BMJ Open Is Committed to Open Peer Review. As Part of This Commitment We Make the Peer Review History of Every Article We Publish Publicly Available

BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected] http://bmjopen.bmj.com/ on September 26, 2021 by guest. Protected copyright. BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Rationale and Design of DAbigatran for Stroke PreVention In Atrial Fibrillation in MoDerate or Severe Mitral Stenosis (DAVID-MS) study ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2020-038194

Article Type: Protocol

Date Submitted by the 04-Mar-2020 Author:

Complete List of Authors: zhou, Mi; University of Hong Kong, Chan, Esther; University of Hong Kong Hai, Jo Jo; The University of Hong Kong, Wong, Chun Ka; The University of Hong Kong, Cardiology Divison, Department of Medicine LAU, Yuk-Ming; University of Hong Kong Huang, Duo; The University of Hong Kong, LAM, Cheung-Chi; University of Hong Kong TAM, Chor-Cheung; University of Hong Kong WONG, Anthony; Queen Mary Hospital, University of Hong Kong, Medicine YUNG, Arthur; Queen Mary Hospital, Univeristy of Hong Kong, Medicine CHAN, Kelvin; Queen Mary Hospital, Univeristy of Hong Kong, Medicine http://bmjopen.bmj.com/ Feng, Yingqing Tan, Ning; Guangdong Cardiovascular Institute, Guangdong provincial Key Laboratory of Coronary Heart Disease Prevention, Guangdong General Hospital, Guangdong Academy of Medical Sciences, Cardiology Chen, Ji-yan; Guangdong Cardiovascular Institute; South China University of Technology YUNG, Chi-Yui; Ruttonjee and Tang Siu Kin Hospital LEE, Kwok-Lun ; Ruttonjee and Tang Siu Kin Hospital , M&G CHOI, Chun-Wai; Tuen Mun Hospital LAM, Ho ; Tuen Mun Hospital, Medicine and Geriatrics on September 26, 2021 by guest. Protected copyright. NG, Andrew; Grantham Hospital FAN, Katherine; Grantham Hospital JIM, Man-Hong; Grantham Hospital Kai Hang, Yiu; Medicine Yan, BP ; Chinese University of Hong Kong, Medicine & Therapeutics SIU, Chung-Wah; The University of Hong Kong,

Valvular heart disease < CARDIOLOGY, CLINICAL PHARMACOLOGY, Keywords: Adult cardiology < CARDIOLOGY

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 25 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

1 2 3 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the terms applicable for US Federal Government officers or employees acting as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author Forunless you peer are acting as review an employee on behalf only of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 25 BMJ Open

1 2 3 Rationale and Design of BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 DAbigatran for Stroke PreVention In Atrial Fibrillation in MoDerate or Severe 7 8 Mitral Stenosis (DAVID-MS) study 9 10 1Mi ZHOU, MBBS; 2Esther W. CHAN, PhD; 1Jojo HAI MBBS; 1Chun-Ka 11 WONG, MBBS; 1Yuk-Ming LAU, MBBS; 1Duo HUANG, MBBS, PhD; 12 1Cheung-Chi LAM, MBBS; 1Chor-Cheung TAM, MBBS; 1 Anthony YT WONG, 13 1 1 3 14 MBBS; Arthur SY YUNG, MBBS; Kelvin KW CHAN, MBBS; Yingqing 15 FENG, MD; 3Ning TAN, MD; 3Ji-Yan CHEN, MD; 4Chi-Yui YUNG, MBBS; 16 4Kwok-Lun LEE, MBBS; 5Chun-Wai CHOI, MB ChB; 5Ho LAM, MBBS; 17 Andrew NG, MBBS; 6Katherine FAN, MBBS; 6Man-Hong JIM, MD; Kai-Hang 18 YIU, MD,For PhD; 1 peer7Bryan P. YANreview MD;# and 1 Chung-Wahonly SIU, MD.# 19 20 1Cardiology Division of Department of Medicine, Queen Mary Hospital, the University 21 2 22 of Hong Kong, Hong Kong SAR, China; Centre for Safe Medication Practice and 23 Research, Department of Pharmacology and Pharmacy, Li Ka Shing Faculty of 3 24 Medicine, the University of Hong Kong, Hong Kong SAR, China; Department of 25 Cardiology, Guangdong Cardiovascular Institute, Guangdong Provincial Key 26 Laboratory of Coronary Heart Disease Prevention, Guangdong General Hospital, 27 Guangdong Academic of Medical Sciences, Guangzhou, China; 4Department of 28 Cardiology, Ruttonjee and Tang Siu Kin Hospital, Hong Kong SAR, China; 29 5Department of Medicine and Geriatrics, Tuen Mun Hospital, Hong Kong SAR, China; 30 6Cardiac Medical Unit, The Grantham Hospital, Hong Kong SAR, China; and 7Division 31 of Cardiology, Department of Medicine and Therapeutics, The Chinese University of 32 Hong Kong, Hong Kong SAR, China. 33 34 #These authors are co-corresponding authors. 35 36 37 Cover title: Dabigatran for Stroke Prevention in Mitral Stenosis and Atrial http://bmjopen.bmj.com/ 38 39 Fibrillation 40 41 42 Declarations of interest: none 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 Correspondence: 48 49 50 Chung-Wah SIU, MD 51 52 Cardiology Division, Department of Medicine, The University of Hong Kong, 53 54 K19, Queen Mary Hospital,102 Pok Fu Lam Road, Hong Kong, China. 55 Tel: (852) 2255-4694, Fax: (852) 2818-6304, 56 57 E-mail: [email protected] & [email protected]. 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 25

1 2

3 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 ABSTRACT 5 6 7 8 Introduction Current international guidelines recommend non-vitamin K oral 9 10 anticoagulants (NOACs) for stroke prevention amongst patients with non- 11 12 valvular atrial fibrillation (AF) at significant ischemic stroke risk given the 13 14 superior safety and comparable efficacy of NOACs over warfarin. Nonetheless, 15 16 17 the safety and effectiveness of NOACs have not been evaluated in patients with 18 For peer review only 19 AF with underlying moderate or severe mitral stenosis (MS), hence the 20 21 recommended stroke prevention strategy remains warfarin therapy. 22 23 24 25 26 Methods and analysis MS remains disproportionately prevalent in Asian 27 28 countries compared with the developed countries. This prospective, 29 30 31 randomized, open-label trial with blinded end-point adjudication aims to 32 33 evaluate the safety and efficacy of dabigatran for stroke prevention in AF 34 35 patients with moderate or severe MS. Patients with AF aged ≥18 years with 36 37 moderate or severe mitral stenosis not planned for valvular intervention in the http://bmjopen.bmj.com/ 38 39 40 coming 12 months will be randomized in 1:1 ratio to receive dabigatran 110 mg 41 42 or 150 mg twice daily or warfarin with INR 2-3 in an open-label design. Patients 43 44 with estimated creatinine clearance <30 ml/min, or with concomitant indication 45 on September 26, 2021 by guest. Protected copyright. 46 47 for anti-platelet therapy will be excluded. The primary outcome is a composite 48 49 of stroke and systemic embolism. Secondary outcomes are ischemic stroke, 50 51 systemic embolism, hemorrhagic stroke, intracranial haemorrhage, major 52 53 54 bleeding and death. The estimated required sample size is approximately 686 55 56 participants. 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 25 BMJ Open

1 2 3 Ethics and dissemination The study protocol of DAVID-MS has been BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 approved by the Institutional Review Board of The University of Hong Kong, 7 8 and Hong Kong West Cluster, Hospital Authority, Hong Kong. 9 10 11 12 PROSPERO registration number 13 14 15 The study is registered with the www.ClinicalTrials.gov (NCT04045093). 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 25

1 2 3 Strengths and limitations of this study BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 This study is the first study to test an alternative to warfarin in patients with AF 7 8 and moderate or severe MS. 9 10 The results will provide important insights to the stroke prevention strategy for 11 12 patients with MS and may be immediately translatable to real clinical practice. 13 14 15 This study will provide the necessary evidence for establishing international 16 17 clinical practice guidelines for stroke prevention in patients with AF and MS. 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 25 BMJ Open

1 2 3 INTRODUCTION BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 7 8 Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia 9 10 encountered in clinical practice.1, 2 Patients with AF are at increased risk of 11 12 ischemic stroke and systemic thromboembolism due to the formation and 13 14 15 embolism of left atrial thrombi,1-3 hence long-term oral anticoagulant (OAC) for 16 17 thromboprophylaxis is the cornerstone in AF management. In previous 18 For peer review only 19 randomized clinical trials in the last century, warfarin has been shown to be 20 21 22 highly effective in reducing stroke risk compared with placebo by as much as 23 24 64% in patients with AF.4 More recently, non-vitamin K oral anticoagulants 25 26 (NOACs) have consistently demonstrated to be safer and more effective for 27 28 29 stroke prevention in patients with non-valvular AF compared with warfarin and 30 31 have become the recommended standard of care for the management of stroke 32 33 prevention in non-valvular AF. 34 35 36 37 http://bmjopen.bmj.com/ 38 While the stroke risk amongst patients with AF appears heterogeneous,5, 6 39 40 patients with underlying valvular heart disease, particularly mitral stenosis (MS) 41 42 are at very high risk for stroke with an annual stroke risk ranging from 4% to 43 44 7 45 17% if left un-anticoagulated. However, patients with AF and underlying MS on September 26, 2021 by guest. Protected copyright. 46 47 are typically excluded in randomized control trials.8 As a result, current 48 49 international guidelines for management of AF do not recommend NOACs for 50 51 9 52 stroke prevention in patients with AF and underlying moderate or severe MS. 53 54 Nonetheless, off-label use of NOAC in patients with AF and MS is not 55 56 uncommon in the real world practice. In a recently published retrospective, 57 58 observational analysis from the Republic of Korea,10 in a cohort of 7,357 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 25

1 2 3 patients with MS receiving anticoagulation therapy, 35% of these patients were BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 in fact treated with NOAC with the remaining 65% with warfarin. More 7 8 importantly, after propensity matching, it was shown that patients treated with 9 10 NOAC had a substantially lower risk of ischemic stroke/systemic embolism with 11 12 an annualized risk of 2.22%/year, compared to that of 4.19%/year for patients 13 14 15 treated with warfarin, (adjusted HR: 0.28; 95% confidence interval (CI): 0.18 to 16 17 0.45), suggesting a potential role of NOAC amongst patients with AF and 18 For peer review only 19 underlying MS.10, 11 20 21 22 23 24 This is of particular importance for Asian AF patients, in whom MS remains 25 26 relatively prevalent despite a declining trend.7 More importantly, the much 27 28 29 higher baseline risk of intracranial haemorrhage and apparently higher 30 31 ischemic stroke risk in Asian populations potentially undermines the benefits of 32 33 warfarin therapy.12,13, 14 Notably, compared with warfarin, the effectiveness and 34 35 safety of NOACs appear to be even more superior in Asian populations than 36 37 http://bmjopen.bmj.com/ 38 Caucasian populations as shown in sub-analyses of pivotal randomized 39 40 controls trials15-17 as well as in studies using real world data.18-23 To our 41 42 knowledge, this is the first multicentre randomized control trial comparing 43 44 45 NOAC to warfarin to address the knowledge in stroke prevention strategy in on September 26, 2021 by guest. Protected copyright. 46 47 patients with AF and moderate or severe MS. This will have immediate and 48 49 long-term impacts on the management of these very high-risk patients with AF. 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 25 BMJ Open

1 2 3 METHODS AND ANALYSIS BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 Study Design 7 8 This is an investigator-initiated, open-label, randomized clinical trial to compare 9 10 effectiveness and safety of dabigatran 150mg or 110 mg twice daily according 11 12 to kidney function with warfarin therapy with target international normalized 13 14 15 ratio (INR) 2-3 for stroke prevention in patients with AF and moderate or severe 16 17 MS. The study is registered with the www.ClinicalTrials.gov (NCT04045093). 18 For peer review only 19 The investigation conforms with the principles outlined in the Declaration of 20 21 22 Helsinki. The study protocol has been approved by the Institutional Review 23 24 Board of The University of Hong Kong, and Hong Kong West Cluster, Hospital 25 26 Authority, Hong Kong. Approvals at other participating sites will be 27 28 29 subsequently obtained. This research protocol complies with the Declaration of 30 31 Helsinki and the International Conference on Harmonisation-Good Clinical 32 33 Practice. 34 35 36 37 http://bmjopen.bmj.com/ 38 Study Participants 39 40 Patients will be recruited from participating specialist cardiology centres in 41 42 Hong Kong SAR China and Mainland China. Written informed consent will be 43 44 45 obtained from all study participants. Table 1 summarizes the inclusion and on September 26, 2021 by guest. Protected copyright. 46 47 exclusion criteria for the study. In brief, patients aged 18 years or above will be 48 49 eligible if they have AF documented on standard 12-lead electrocardiography 50 51 52 (ECG) performed at screening or randomization and moderate or severe MS 53 54 as defined as the mitral valvular area (MVA) of 1.0-1.5cm2 and <1.0 cm2, 55 56 respectively. Reasons for exclusion include the presence of prosthetic valve, 57 58 left atrial appendage occlusive device, and/or active endocarditis; planned 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 25

1 2 3 valvular intervention and/or planned AF ablation; history of major bleeding BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 including intracranial, intraocular, spinal or retroperitoneal haemorrhage; 7 8 unexplained anaemia with haemoglobin level <10 g/dL, or thrombocytopenia 9 10 with platelet count <100×109/L; need for anticoagulant or antiplatelet therapy of 11 12 conditions other than AF; concomitant use of potent P-gp inhibitor(s) or drugs 13 14 15 with known interaction with dabigatran; uncontrolled hypertension; significant 16 17 kidney impairment with estimated creatinine clearance (CrCl) ≤30 mL/min; liver 18 For peer review only 19 dysfunction of Child Pugh Stage B or C; pregnancy or if there is child-baring 20 21 22 potential during the full duration of the study. In addition, patients considered 23 24 unsuitable by the investigator including short life expectancy <1 year due to 25 26 concomitant disease, substance and/or alcohol abuse or other medical 27 28 29 conditions. 30 31 32 33 Study Procedures 34 35 After providing written informed consent, all study participants will be randomly 36 37 http://bmjopen.bmj.com/ 38 assigned to receive dabigatran or to receive warfarin. The procedure of the trial 39 40 is summarized in figure 1, For patients randomized to receive dabigatran, the 41 42 dosage regimen will be determined according to the respective estimated CrCl 43 44 45 or if concomitantly taking interacting drugs requiring dosage adjustment. on September 26, 2021 by guest. Protected copyright. 46 47 Patients with estimated CrCl above 50 ml/min will receive dabigatran 150 mg 48 49 twice daily, whereas those with CrCl between 30 to 50 ml/min will receive 50 51 52 dabigatran 110 mg twice daily. For patients previously on warfarin randomized 53 54 to receive dabigatran, dabigatran will initiate after discontinuation of warfarin 55 56 with an INR less than or equal to 2. At the end of study, patients randomized to 57 58 dabigatran will be switched back to warfarin. Warfarin will be initiated 3 days 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 25 BMJ Open

1 2 3 prior to the termination of dabigatran. INR will be checked 5 days after initiation BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 of warfarin i.e., 2 days after termination of dabigatran to minimise potential 7 8 impact of remaining dabigatran levels in elevating the INR. On the other hand, 9 10 for those randomized to receive warfarin, INR will be measured at least every 11 12 8 weeks with target INR of 2.0 to 3.0. The time in therapeutic range (TTR) will 13 14 15 be calculated for each study participant using Rosendaal method,24 in which 16 17 INR will be assumed to change in a linear manner between measurements, and 18 For peer review only 19 INR values on the days without measurement are interpolated. The percentage 20 21 22 of time during which a study participant has an INR within 2.0-3.0 is taken as 23 24 TTR. The first follow-up visit will be scheduled 14 days after randomization and 25 26 then every 4 months during the study period of 1 year (Table 2). 27 28 29 30 31 Outcomes 32 33 The primary outcome is a composite of stroke or systemic embolism at 1 year. 34 35 Secondary outcomes are ischemic stroke, systemic embolism, hemorrhagic 36 37 http://bmjopen.bmj.com/ 38 stroke, intracranial haemorrhage, major bleeding and death at 1 year. Stroke is 39 40 defined as a neurological deficit of sudden onset that persisted for more than 41 42 24 hours and corresponded to a vascular territory that cannot be explained by 43 44 45 other causes (trauma, infection, vasculitis). Stroke will be further classified as on September 26, 2021 by guest. Protected copyright. 46 47 ischemic stroke and hemorrhagic stroke according to computerized axial 48 49 tomography or magnetic resonance imaging of the brain. Intracranial 50 51 52 hemorrhage (ICH) consists of hemorrhagic stroke (intra-cerebral hemorrhage 53 54 and cerebellar hemorrhage), subdural hemorrhage, and subarachnoid 55 56 hemorrhage, and will be confirmed with computerized axial tomography or 57 58 magnetic resonance imaging of the brain. Systemic embolism is defined as an 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 25

1 2 3 acute vascular occlusion of an extremity or organ other than the brain, BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 documented by imaging, surgery, and/or autopsy. 7 8 9 10 Major bleeding is defined as a drop in the hemoglobin level of at least 2 g/dL, 11 12 transfusion of at least 2 units of blood, or symptomatic bleeding in a critical area 13 14 15 or organ. Life-threatening bleeding includes fatal bleeding, symptomatic 16 17 intracranial bleeding, bleeding with a hemoglobin drop of at least 5 g/dL, or 18 For peer review only 19 bleeding requiring transfusion of at least 4 units of blood or inotropic agents or 20 21 22 requiring surgery. All outcomes will be adjudicated by 2 independent 23 24 investigators in a blinded fashion. 25 26 27 28 29 Randomization 30 31 Randomization will be stratified to each study site to account for variations in 32 33 patient demographics and diagnoses. At each site, patients will be randomised 34 35 to “permuted blocks of four” (two of each study arm) to assist in equality of 36 37 http://bmjopen.bmj.com/ 38 numbers in each arm. An independent research officer will generate the 39 40 random-number table. The codes will remain securely in the Department of 41 42 Pharmacy, QMH until study completion and analysis of results. 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 Patient monitoring and safety 48 49 An independent Safety Committee will be established comprising of an 50 51 52 Emergency Clinician, Clinical Pharmacologist and Toxicologist. They will 53 54 receive regular reports during patient enrolment and be notified of any adverse 55 56 drug reaction and study protocol violation. The Safety Committee is led by 57 58 Professor Bernard Cheung from the Clinical Pharmacology, Faculty of 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 25 BMJ Open

1 2 3 Medicine, the University of Hong Kong, Hong Kong. For patient safety, BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 discontinuation of the study is at the discretion of the cardiology clinician to 7 8 enable informed decisions to be made regarding subsequent management and 9 10 alternative medication use. Any medication or therapy, intervention or 11 12 procedure thought to be necessary for the safe management of the patient may 13 14 15 be administered at the discretion of the managing clinician. 16 17 18 For peer review only 19 Sample Size Calculation 20 21 22 The primary analysis is to test whether dabigatran is noninferior to warfarin for 23 24 ischemic stroke prevention in patients with AF and moderate or severe MS. The 25 26 potential for dabigatran to preserve at least 50% of the effectiveness of warfarin 27 28 29 is considered clinically meaningful, as noninferior in patients with AF and 30 31 moderate or severe MS. The noninferiority margin is 1.49, which is derived from 32 33 the only observation study comparing vitamin K antagonist with NOAC in 34 35 patients with AF and MS.10 In the study, the annual ischemic stroke risk of 36 37 http://bmjopen.bmj.com/ 38 patients with AF and MS receiving vitamin K antagonist and NOAC are 39 40 4.19%/year and 2.22%/year respectively. Accordingly, a sample size of 686 41 42 patients (343 patients in the vitamin K antagonist group and 343 in the 43 44 45 dabigatran group) including 10% attrition would be needed to satisfy the on September 26, 2021 by guest. Protected copyright. 46 47 noninferiority hypothesis with the upper boundary of the one-sided 95% 48 49 confidence interval (CI) (or equivalent with a 90% two-sided CI) and the Hazard 50 51 52 ratio (HR) of the primary outcome below the noninferiority margin of 1.49. 53 54 55 56 Patient and public involvement 57 58 No patient and public involved in the research plan of this study. 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 25

1 2 3 Statistical Analysis BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 Baseline data will be reported as means and standard deviations for continuous 7 8 data and as numbers and percentages for categorical data. All endpoints will 9 10 be analysed according to the intention-to-treat principle, with all patients who 11 12 undergo randomization included in the analysis. Clinical events that occur after 13 14 15 randomization and until the end of the study (at 1 year or mortality) will be 16 17 included in the primary analysis of clinical outcomes. A p-value <0.05 18 For peer review only 19 considered as significant. Calculations will be performed using SPSS software 20 21 22 (version 12.0). 23 24 25 26 Storage and Security 27 28 29 To minimise lost or misplaced data collection sheets, each document will 30 31 contain clear contact information for each study site. An Excel file containing 32 33 study data will be kept in de-identified form for 5 years. Data will be kept on a 34 35 password-protected data files on a local computer drive at the study sites, 36 37 http://bmjopen.bmj.com/ 38 accessible only by the study investigators. Back-up discs will be encrypted and 39 40 kept locked securely. All paper records will be de-identified and stored securely 41 42 in a locked cabinet for 5 years. Presented and published data will not allow 43 44 45 identification of any study subject. on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 25 BMJ Open

1 2 3 DISCUSSION BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 7 8 To our knowledge, this is the first study to evaluate dabigatran as an alternative 9 10 to warfarin for stroke prevention in patients with AF and moderate or severe 11 12 MS. The results could fill the gap in stroke prevention strategy for this specific 13 14 15 group of patients with AF with immediate and long-term impacts on clinical 16 17 practice. 18 For peer review only 19 20 21 22 AF is the most commonly encountered sustained cardiac arrhythmia in clinical 23 24 practice.1, 2 While patients with AF have in general an increased risk of stroke, 25 26 4 subgroups are at particularly high risk necessitating long-term anticoagulation 27 28 29 therapy. These include (1) patients with non-valvular AF and high CHA2DS2- 30 31 VASc score, (2) patients with AF and hypertrophic cardiomyopathy, (3) patients 32 33 with AF and MS, and (4) patients with AF and mechanical heart valves. (Figure 34 35 2) In the past decade, NOAC has emerged as the preferred agent over warfarin 36 37 http://bmjopen.bmj.com/ 38 for stroke prevention in patients with non-valvular AF and high CHA2DS2-VASc 39 40 score. In the 4 pivotal studies comparing NOACs and warfarin in patients with 41 42 non-valvular AF, NOACs are at least as effective as warfarin to reduce stroke 43 44 45 and at the same time are much safer alternatives in term of life-threatening on September 26, 2021 by guest. Protected copyright. 46 47 bleeding complications. Nonetheless, direct extrapolation of these results to 48 49 other subgroups of AF patients may not be appropriate, given the different 50 51 52 mechanisms of thrombus formation in different diseases. For instance, in the 53 54 RE-ALIGN trial randomizing 252 patients with recent mechanical valvular 55 56 replacement in a 2:1 ratio to receive either dabigatran or warfarin, there was an 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 25

1 2 3 excess of thromboembolic as well as bleeding events among patients BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 25 6 randomized to dabigatran, rendering the study prematurely terminated. 7 8 9 10 On the other hand, international guidelines do not recommend NOACs for 11 12 patients with AF and moderate or severe mitral stenosis due to the lack of 13 14 15 reliable data from clinical trials. Nonetheless, it remains undetermined whether 16 17 NOACs can be used as an alternative to warfarin for patients with AF and 18 For peer review only 19 moderate or severe MS due to the lack of clinical trial. The current study has 20 21 22 several important implications particularly in Asian countries. First, while MS is 23 24 now a rare condition in developed countries, it remains relatively prevalent in 25 26 many Asian countries. In addition, the risk of stroke amongst patients with AF 27 28 29 and MS is only second to those with mechanical valvular replacement ranging 30 31 from 4 to 17%. Second, previous epidemiological studies12,13, 14, 21, 26, 27 and 32 33 sub-analyses of the pivotal NOAC trials15-17 have consistently reported a much 34 35 higher nominal risk of ICH amongst Asians than non-Asians, favouring NOACs 36 37 http://bmjopen.bmj.com/ 38 over warfarin therapy. More importantly, the notoriously poor time in therapeutic 39 40 range (TTR) for warfarin in Asian populations observed in real world data18, 27- 41 42 31 and pivotal NOAC trials15-17 substantially undermines the overall clinical 43 44 45 benefits of warfarin therapy. In fact, the annual incidence of ICH amongst on September 26, 2021 by guest. Protected copyright. 46 47 patients with AF and MS treated with warfarin has been reported to be as high 48 49 as 0.93% per year,10 urging a much safer alternative. 50 51 52 53 54 In the present study, the NOAC of choice is dabigatran, the first NOAC with 55 56 approved indication for stroke prevention in non-valvular AF patients from the 57 58 United States Food and Drug Administration in 2009. In the pivotal study, the 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 25 BMJ Open

1 2 3 RE-LY study,32 patients with non-valvular AF with CHADS2>1 were randomly BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 assigned to two doses of dabigatran: 110 mg or 150 mg twice daily; or adjusted- 7 8 dose warfarin. After a median follow-up of 2.0 years, the low-dose regime was 9 10 found to be as effective as warfarin in preventing the primary endpoint (a 11 12 composite of stroke and systemic embolism) (1.52%/year vs. 1.69%/year), but 13 14 15 with substantially lower risk of major bleeding and ICH.32 On the other hand, 16 17 the standard dose dabigatran (150mg BD) is superior to warfarin in reducing 18 For peer review only 19 the primary composite endpoint and ICH, with a comparable risk of major 20 21 32 22 bleeding. In an analysis comparing the effectiveness and safety of dabigatran 23 24 according to the ethnicity of study participants, dabigatran appears to be more 25 26 effective in stroke prevention as well as safer in terms of ICH compared with 27 28 29 warfarin. This is in concordance to subsequent real-world cohorts of AF patients 30 31 from territory-wide registries from Asia Pacific region. Plausible explanations 32 33 include suboptimal quality of warfarin therapy with low time in therapeutic range 34 35 and higher risk of ICH in Asian populations.18, 23, 28, 33, 34 An additional reason 36 37 http://bmjopen.bmj.com/ 38 for the choice of dabigatran in the present study is the wide availability of its 39 40 antidote, idarucizumab in Asian countries, which provides extra-protection of 41 42 patients in the clinical trial. 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 Conclusion 48 49 50 The study is designed to provide clinicians with robust, much ‐ needed 51 52 53 information regarding stroke prevention strategy for patients with AF and 54 55 moderate or severe MS. The results will have immediate and long-term impacts 56 57 on the management of these very high-risk patients with AF. 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 25

1 2 3 Contributors BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 Mi ZHOU is the investigator of a local site and is responsible for the study 7 8 design, study dataset and website establishment, and study execution. 9 10 Esther W. CHAN, Jojo HAI, Chun-Ka WONG, Yuk-Ming LAU, Duo HUANG, 11 12 Cheung-Chi LAM, Chor-Cheung TAM, Yiu-Tung WONG, See-Yue YUNG, Ki- 13 14 15 Wan CHAN, Yingqing FENG, Ning TAN, Ji-Yan CHEN, Chi-Yui YUNG, Kwok- 16 17 Lun LEE, Chun-Wai CHOI, Ho LAM, Andrew NG, Katherine FAN, Man-Hong 18 For peer review only 19 JIM, Kai-Hang YIU, those people above are the investigator of a local site. 20 21 22 Bryan P. YAN is the principal investigator of the whole study responsible for 23 24 study design. 25 26 Chung-Wah SIU is the principal investigator of the whole study responsible for 27 28 29 study design, study execution, manuscript drafting, and study site recruitment. 30 31 Acknowledgement: None. 32 33 Funding: None. 34 35 Conflict of Interest: None. 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 25 BMJ Open

1 2 3 LEGENDS: BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 Figure 1. Design for the DAVID-MS study. 7 8 9 10 Figure 2. Four main groups of patients with AF requiring long-term 11 12 anticoagulation therapy. MS: mitral stenosis; NOAC: non-vitamin K oral 13 14 15 anticoagulant; and VKA: vitamin K antagonist. 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 25

1 2 3 Table 1. Inclusion and Exclusion Criteria BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 Inclusion Criteria 7 8  Patients with AF documented with standard 12-lead ECG documented AF 9 on the day of screening or randomization 10 11  Patients with age >18 years

12 2 13  Patients with moderate or severe MS i.e., MVA <1.5 cm 14  Patients should be able to provide a written, informed consent. 15 16  Patients should have all 4 inclusion-criteria fulfilled to be qualified for the 17 study. 18 For peer review only 19 Exclusion criteria 20 21  Patients with prosthetic valve, or with active endocarditis 22 Patients with planned valvular intervention within 1 year 23  24  Patients with left atrial appendage occlusive device 25 26  Patients with planned AF ablation 27  Patients with history of intracranial, intraocular, spinal, or retroperitoneal 28 29 bleeding 30 31  Unexplained anemia (hemoglobin level <10 g/dL) or thrombocytopenia 32 (platelet count <100×109/L) 33 34  Need for anticoagulant therapy of disorders other than AF 35 36  Patients receiving antiplatelet therapy for disorders other than AF 37  Patients receiving concomitant P-gp inhibitors and/or medications known to http://bmjopen.bmj.com/ 38 39 interact with dabigatran 40  Uncontrolled hypertension (systolic blood pressure >180 mmHg and/or 41 42 diastolic blood pressure >100 mmHg) 43 44  Estimated creatinine clearance ≤30 mL/min 45  Liver dysfunction of Child Pugh stage B or C on September 26, 2021 by guest. Protected copyright. 46 47  Women who are pregnant or of childbearing potential who refuse to use a 48 medically acceptable form of contraception throughout the study 49 50  Patients considered unreliable by the investigator or have a life expectancy 51 52 less than 1 year because of concomitant disease, or has any condition, 53 which in the opinion of the investigator, would not allow safe participation in 54 55 the study (e.g. drug addiction, alcohol abuse). 56 57 Abbreviations: AF: atrial fibrillation; MS: mitral stenosis; MVA: mitral valvular 58 area 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Page 21 of 25 BMJ Open

1 2 3 4 5 Table 2. Study visits 6 Visits -1 0 1 2 3 4 5 6 7 UNS** EOS*** 7 8 Weeks -2 to 0 0 8 16 24 32 40 48 56 -- 56-60 9 10 Informed consent X 11 12 Inclusion & exclusion criteriaFor peerX review only 13 14 15 Randomization X 16 17 Medical history X http://bmjopen.bmj.com/ 18 19 Physical examination X X X X X X X X X X 20 21 22 Echocardiography X 23 24 INR X X* X* X* X* X* X* X* X* X 25 on September 26, 2021 by guest. Protected copyright. 26 Renal function X X X X X X 27 28 29 Drug dispensing X X X X X X X X 30 31 Drug collection X X X X X X X X 32 33 Outcome events X X X X X X X X X 34 35 36 Adverse events X X X X X X X X X 37 38 39 40 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open Page 22 of 25

1 2 3 4 * 5 Only for patients randomized to receive warfarin 6 **UNS (unplanned visit): (X) The marked item is optimal and performed according to the judgment of researchers. 7 ***EOS (final visit): Make arrangement according to the study end time (if there is a visit within one month before the end of study, it 8 9 is regarded as a final visit, but needs to be supplemented with the items required completely) 10 11 12 For peer review only 13 14 15 16

17 http://bmjopen.bmj.com/ 18 19 20 21 22 23 24

25 on September 26, 2021 by guest. Protected copyright. 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 23 of 25 BMJ Open

1 2

3 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 References: 5 6 7 1. Camm AJ, Kirchhof P, Lip GY, Schotten U, Savelieva I, Ernst S, et al. 8 Guidelines for the management of atrial fibrillation: the Task Force for 9 the Management of Atrial Fibrillation of the European Society of 10 Cardiology (ESC). Europace 2010;12(10):1360-420. 11 2. Lip GY, Tse HF, Lane DA. Atrial fibrillation. Lancet 12 2012;379(9816):648-61. 13 14 3. Camm AJ, Lip GY, De Caterina R, Savelieva I, Atar D, Hohnloser SH, 15 et al. 2012 focused update of the ESC Guidelines for the management 16 of atrial fibrillation: an update of the 2010 ESC Guidelines for the 17 management of atrial fibrillation--developed with the special 18 contributionFor of peer the European review Heart Rhythm only Association. Europace : 19 European pacing, arrhythmias, and cardiac electrophysiology : journal 20 21 of the working groups on cardiac pacing, arrhythmias, and cardiac 22 cellular electrophysiology of the European Society of Cardiology 23 2012;14(10):1385-413. 24 4. Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotic therapy 25 to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann 26 Intern Med 2007;146(12):857-67. 27 5. Siu CW, Pong V, Zhang X, Chan YH, Jim MH, Liu S, et al. Risk of 28 29 ischemic stroke after new-onset atrial fibrillation in patients with 30 hyperthyroidism. Heart Rhythm 2009;6(2):169-73. 31 6. Guo Y, Wang H, Tian Y, Wang Y, Lip GY. Multiple risk factors and 32 ischaemic stroke in the elderly Asian population with and without atrial 33 fibrillation. An analysis of 425,600 Chinese individuals without prior 34 stroke. Thromb Haemost 2016;115(1):184-92. 35 7. Chandrashekhar Y, Westaby S, Narula J. Mitral stenosis. Lancet 36 37 2009;374(9697):1271-83. http://bmjopen.bmj.com/ 38 8. Breithardt G, Baumgartner H, Berkowitz SD, Hellkamp AS, Piccini JP, 39 Stevens SR, et al. Clinical characteristics and outcomes with 40 rivaroxaban vs. warfarin in patients with non-valvular atrial fibrillation 41 but underlying native mitral and aortic valve disease participating in the 42 ROCKET AF trial. Eur Heart J 2014;35(47):3377-85. 43 44 9. Kirchhof P, Benussi S, Kotecha D, Ahlsson A, Atar D, Casadei B, et al. 45 2016 ESC Guidelines for the management of atrial fibrillation on September 26, 2021 by guest. Protected copyright. 46 developed in collaboration with EACTS. Europace 2016;18(11):1609- 47 1678. 48 10. Kim JY, Kim SH, Myong JP, Kim YR, Kim TS, Kim JH, et al. Outcomes 49 of Direct Oral Anticoagulants in Patients With Mitral Stenosis. J Am 50 Coll Cardiol 2019;73(10):1123-1131. 51 52 11. Giugliano RP, O'Gara PT. DOACs in Patients With Mitral Stenosis and 53 Atrial Fibrillation: Time for a Randomized Clinical Trial. J Am Coll 54 Cardiol 2019;73(10):1132-1134. 55 12. Tse HF, Wang YJ, Ai-Abdullah MA, Pizarro-Borromeo AB, Chiang CE, 56 Krittayaphong R, et al. Stroke Prevention in Atrial Fibrillation - An Asian 57 Stroke Perspective. Heart Rhythm 2013. 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 25

1 2 3 13. Chong BH, Chan KH, Pong V, Lau KK, Chan YH, Zuo ML, et al. Use of BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 aspirin in Chinese after recovery from primary intracranial 5 6 haemorrhage. Thromb Haemost 2012;107(2):241-7. 7 14. Friberg L, Rosenqvist M, Lip GY. Evaluation of risk stratification 8 schemes for ischaemic stroke and bleeding in 182 678 patients with 9 atrial fibrillation: the Swedish Atrial Fibrillation cohort study. Eur Heart J 10 2012;33(12):1500-10. 11 15. Hori M, Connolly SJ, Zhu J, Liu LS, Lau CP, Pais P, et al. Efficacy and 12 safety of dabigatran versus warfarin in patients with atrial fibrillation: 13 14 analysis in Asian population in RE-LY trial. Cerebrovascular Disease 15 2012;34 (Suppl 1)(Asia Pacific Stroke Conference 2012):9. 16 16. Hankey GJ, Stevens S, Piccini JP, Lokhnygina Y, Mahaffey KW, 17 Halperin JL, et al. Predictors of Intracranial Hemorrhage among 18 anticoagulatedFor peer patients review with atrial fibrillation: only insights from the 19 Rivaroxaban once daily oral direct factor Xa inhibition compared 20 21 with vitamin K antagonism for prevention of stroke and embolism trial in atrial 22 fibrillation (ROCKET AF). Stroke 2012;43:A152. 23 17. Tanahashi N, Hori M, Matsumoto M, Momomura SI, Uchiyama S, Goto 24 S, et al. Rivaroxaban versus Warfarin in Japanese Patients with 25 Nonvalvular Atrial Fibrillation for the Secondary Prevention of Stroke: A 26 Subgroup Analysis of J-ROCKET AF. J Stroke Cerebrovasc Dis 2013. 27 18. Chan PH, Huang D, Lau CP, Chan EW, Wong IC, Lip GY, et al. Net 28 29 Clinical Benefit of Dabigatran Over Warfarin in Patients With Atrial 30 Fibrillation Stratified by CHA2DS2-VASc and Time in Therapeutic 31 Range. Can J Cardiol 2016;32(10):1247 e15-1247 e21. 32 19. Chan PH, Li WH, Hai JJ, Chan KH, Tse HF, Cheung BM, et al. 33 Gastrointestinal haemorrhage in atrial fibrillation patients: impact of 34 quality of anticoagulation control. Eur Heart J Cardiovasc 35 Pharmacother 2015;1(4):265-72. 36 37 20. Lee YK, Lau YM, Cai ZJ, Lai WH, Wong LY, Tse HF, et al. Modeling http://bmjopen.bmj.com/ 38 Treatment Response for Lamin A/C Related Dilated Cardiomyopathy in 39 Human Induced Pluripotent Stem Cells. J Am Heart Assoc 2017;6(8). 40 21. Lau WCY, Li X, Wong ICK, Man KKC, Lip GYH, Leung WK, et al. 41 Bleeding-related hospital admissions and 30-day readmissions in 42 patients with non-valvular atrial fibrillation treated with dabigatran 43 44 versus warfarin. J Thromb Haemost 2017;15(10):1923-1933. 45 22. Huang D, Cheng YY, Chan PH, Hai J, Yiu KH, Tse HF, et al. Rationale on September 26, 2021 by guest. Protected copyright. 46 and design of the screening of pulmonary hypertension in systemic 47 lupus erythematosus (SOPHIE) study. ERJ Open Res 2018;4(1). 48 23. Qi X, Wong BL, Lau SH, Ng KT, Kwok SY, Kin-Wai Sun C, et al. A 49 hemoglobin-based oxygen carrier sensitized Cisplatin based 50 chemotherapy in hepatocellular carcinoma. Oncotarget 51 52 2017;8(49):85311-85325. 53 24. Rosendaal FR, Cannegieter SC, van der Meer FJ, Briet E. A method to 54 determine the optimal intensity of oral anticoagulant therapy. Thromb 55 Haemost 1993;69(3):236-9. 56 25. Eikelboom JW, Connolly SJ, Brueckmann M, Granger CB, Kappetein 57 AP, Mack MJ, et al. Dabigatran versus warfarin in patients with 58 mechanical heart valves. N Engl J Med 2013;369(13):1206-14. 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 25 BMJ Open

1 2 3 26. Chan EW, Lau WC, Siu CW, Lip GY, Leung WK, Anand S, et al. Effect BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 of suboptimal anticoagulation treatment with antiplatelet therapy and 5 6 warfarin on clinical outcomes in patients with nonvalvular atrial 7 fibrillation: A population-wide cohort study. Heart Rhythm 8 2016;13(8):1581-8. 9 27. Teo KC, Mahboobani NR, Lee R, Siu CW, Cheung RT, Ho SL, et al. 10 Warfarin associated intracerebral hemorrhage in Hong Kong Chinese. 11 Neurol Res 2014;36(2):143-9. 12 28. Huang D, Wong CL, Cheng KW, Chan PH, Yue WS, Wong CK, et al. 13 14 Impact of provision of time in therapeutic range value on 15 anticoagulation management in atrial fibrillation patients on warfarin. 16 Postgrad Med J 2018;94(1110):207-211. 17 29. Chan PH, Hai JJ, Chan EW, Li WH, Tse HF, Wong IC, et al. Use of the 18 SAMe-TT2R2For peer Score to Predictreview Good Anticoagulation only Control with 19 Warfarin in Chinese Patients with Atrial Fibrillation: Relationship to 20 21 Ischemic Stroke Incidence. PLoS One 2016;11(3):e0150674. 22 30. Ho CW, Ho MH, Chan PH, Hai JJ, Cheung E, Yeung CY, et al. 23 Ischemic stroke and intracranial hemorrhage with aspirin, dabigatran, 24 and warfarin: impact of quality of anticoagulation control. Stroke 25 2015;46(1):23-30. 26 31. Siu CW, Tse HF. Net clinical benefit of warfarin therapy in elderly 27 Chinese patients with atrial fibrillation. Circ Arrhythm Electrophysiol 28 29 2014;7(2):300-6. 30 32. Connolly SJ, Ezekowitz MD, Yusuf S, Eikelboom J, Oldgren J, Parekh 31 A, et al. Dabigatran versus warfarin in patients with atrial fibrillation. N 32 Engl J Med 2009;361(12):1139-51. 33 33. Hai JJ, Chan PH, Chan YH, Fong CH, Huang D, Li WH, et al. 34 Prediction of Thromboembolic Events in Heart Failure Patients in Sinus 35 Rhythm: The Hong Kong Heart Failure Registry. PLoS One 36 37 2016;11(12):e0169095. http://bmjopen.bmj.com/ 38 34. Chan PH, Li WH, Hai JJ, Chan EW, Wong IC, Tse HF, et al. Time in 39 Therapeutic Range and Percentage of International Normalized Ratio 40 in the Therapeutic Range as a Measure of Quality of Anticoagulation 41 Control in Patients With Atrial Fibrillation. Can J Cardiol 42 2016;32(10):1247 e23-1247 e28. 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 25 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Figure 1. Design for the DAVID-MS study

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 25 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 Figure 2. Four main groups of patients with AF requiring long-term anticoagulation therapy. MS: mitral stenosis; NOAC: non-vitamin K oral anticoagulant; and VKA: vitamin K antagonist. 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Protocol for Rationale and Design of DAbigatran for Stroke PreVention In Atrial Fibrillation in MoDerate or Severe Mitral Stenosis (DAVID-MS) study

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2020-038194.R1

Article Type: Protocol

Date Submitted by the 12-May-2020 Author:

YUNG, Arthur; Queen Mary Hospital, Univeristy of Hong Kong, Medicine http://bmjopen.bmj.com/ CHAN, Kelvin; Queen Mary Hospital, Univeristy of Hong Kong, Medicine Feng, Yingqing; Guangdong Cardiovascular Institute Tan, Ning; Guangdong Cardiovascular Institute, Guangdong provincial Key Laboratory of Coronary Heart Disease Prevention, Guangdong General Hospital, Guangdong Academy of Medical Sciences, Cardiology Chen, Ji-yan; Guangdong Cardiovascular Institute; South China University of Technology YUNG, Chi-Yui; Ruttonjee and Tang Siu Kin Hospital LEE, Kwok-Lun ; Ruttonjee and Tang Siu Kin Hospital , M&G on September 26, 2021 by guest. Protected copyright. CHOI, Chun-Wai; Tuen Mun Hospital LAM, Ho ; Tuen Mun Hospital, Medicine and Geriatrics NG, Andrew; Grantham Hospital FAN, Katherine; Grantham Hospital JIM, Man-Hong; Grantham Hospital Kai Hang, Yiu; Medicine Yan, BP ; Chinese University of Hong Kong, Medicine & Therapeutics SIU, Chung-Wah; The University of Hong Kong,

Primary Subject Cardiovascular medicine Heading:

Secondary Subject Heading: Pharmacology and therapeutics

Valvular heart disease < CARDIOLOGY, CLINICAL PHARMACOLOGY, Keywords: Adult cardiology < CARDIOLOGY

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 52 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 52 BMJ Open

1 2 3 Protocol for Rationale and Design of BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 DAbigatran for Stroke PreVention In Atrial Fibrillation in 7 8 MoDerate or Severe Mitral Stenosis (DAVID-MS) study 9 10 11 1Mi ZHOU, MBBS; 2Esther W. CHAN, PhD; 1Jojo HAI MBBS; 1Chun-Ka 12 WONG, MBBS; 1Yuk-Ming LAU, MBBS; 1Duo HUANG, MBBS, PhD; 13 1Cheung-Chi LAM, MBBS; 1Chor-Cheung TAM, MBBS; 1 Anthony YT WONG, 14 1 1 3 15 MBBS; Arthur SY YUNG, MBBS; Kelvin KW CHAN, MBBS; Yingqing 3 3 4 16 FENG, MD; Ning TAN, MD; Ji-Yan CHEN, MD; Chi-Yui YUNG, MBBS; 17 4Kwok-Lun LEE, MBBS; 5Chun-Wai CHOI, MB ChB; 5Ho LAM, MBBS; 18 Andrew NG, MBBS;For 6Katherinepeer FAN,review MBBS; 6Man-Hong only JIM, MD; Kai-Hang 19 YIU, MD, PhD;1 7Bryan P. YAN MD;# and 1Chung-Wah SIU, MD.# 20 21 1Cardiology Division, Department of Medicine, Queen Mary Hospital, the 22 2 23 University of Hong Kong, Hong Kong SAR, China; Centre for Safe Medication 24 Practice and Research, Department of Pharmacology and Pharmacy, Li Ka 25 Shing Faculty of Medicine, the University of Hong Kong, Hong Kong SAR, 26 China; 3Department of Cardiology, Guangdong Cardiovascular Institute, 27 Guangdong Provincial Key Laboratory of Coronary Heart Disease Prevention, 28 Guangdong General Hospital, Guangdong Academic of Medical Sciences, 29 Guangzhou, China; 4Department of Cardiology, Ruttonjee and Tang Siu Kin 30 5 31 Hospital, Hong Kong SAR, China; Department of Medicine and Geriatrics, 6 32 Tuen Mun Hospital, Hong Kong SAR, China; Cardiac Medical Unit, The 33 Grantham Hospital, Hong Kong SAR, China; and 7Division of Cardiology, 34 Department of Medicine and Therapeutics, The Chinese University of Hong 35 Kong, Hong Kong SAR, China. 36 37 # http://bmjopen.bmj.com/ 38 These authors are co-corresponding authors. 39 40 Protocol Version: 1 41 42 Protocol Date: March 10, 2020 43 Wordcounts: 3452 44 Keywords: Mitral stenosis, atrial fibrillation, non-vitamin K oral anticoagulant, 45 on September 26, 2021 by guest. Protected copyright. 46 warfarin 47 48 Correspondence: 49 50 Chung-Wah SIU, MD 51 Cardiology Division, 52 Department of Medicine, 53 The University of Hong Kong, 54 55 Hong Kong SAR, China 56 Tel: (852) 2255-4694, 57 Fax: (852) 2818-6304, 58 59 E-mail: [email protected] & [email protected]. 60

1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 52

1 2 3 ABSTRACT BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 Introduction: Current international guidelines recommend non-vitamin K oral 7 8 9 anticoagulants (NOACs) for stroke prevention amongst patients with non- 10 11 valvular atrial fibrillation (AF) at significant ischaemic stroke risk given the 12 13 superior safety and comparable efficacy of NOACs over warfarin. Nonetheless, 14 15 16 the safety and effectiveness of NOACs have not been evaluated in patients with 17 18 AF with underlyingFor moderatepeer orreview severe mitral only stenosis (MS), hence the 19 20 recommended stroke prevention strategy remains warfarin therapy. 21 22 23 24 25 Method and analysis: MS remain disproportionately prevalent in Asian 26 27 countries compared with the developed countries. This prospective, 28 29 randomized, open-label trial with blinded endpoint adjudication aims to evaluate 30 31 32 the safety and efficacy of dabigatran for stroke prevention in AF patients with 33 34 moderate or severe MS. Patients with AF aged ≥18 years with moderate or 35 36 severe mitral stenosis not planned for valvular intervention in the coming 12 37 http://bmjopen.bmj.com/ 38 39 months will be randomized in a 1:1 ratio to receive dabigatran 110 mg or 150 40 41 mg twice daily or warfarin with INR 2-3 in an open-label design. Patients with 42 43 estimated creatinine clearance <30 ml/min, or with a concomitant indication for 44 45 on September 26, 2021 by guest. Protected copyright. 46 anti-platelet therapy will be excluded. The primary outcome is a composite of 47 48 stroke and systemic embolism. Secondary outcomes are ischaemic stroke, 49 50 systemic embolism, hemorrhagic stroke, intracranial haemorrhage, major 51 52 bleeding, and death. The estimated required sample size is approximately 686 53 54 55 participants. 56 57 58 59 60

2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 52 BMJ Open

1 2 3 Ethics and dissemination: The study protocol of DAVID-MS has been BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 approved by the Institutional Review Board of The University of Hong Kong, 7 8 and Hong Kong West Cluster, Hospital Authority, Hong Kong. Results will be 9 10 11 12 published in peer-reviewed journals. 13 14 15 16 17 Registration details: ClinicalTrials.gov (NCT04045093). 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

3 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 52

1 2 3 STRENGTHS AND LIMITATIONS OF THIS STUDY BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6  This is the first investigator-initiated, open-label, randomized clinical trial 7 8 to compare effectiveness and safety of dabigatran and warfarin therapy 9 10 for stroke prevention in patients with atrial fibrillation (AF) and moderate 11 12 13 or severe mitral stenosis (MS). 14 15  The study is designed to provide clinicians with robust, much-needed 16 17 information regarding stroke prevention strategy for patients with AF and 18 For peer review only 19 20 moderate or severe MS. 21 22  The results of study will have immediate and long-term impacts on the 23 24 management of these very high-risk patients with AF. 25 26 27  This study will provide the necessary evidence for establishing 28 29 international clinical practice guidelines for stroke prevention in patients 30 31 with AF and MS. 32 33 34  Since the clinical trial will be conducted mainly in Hong Kong and 35 36 mainland China, it is expected that most recruited subjects will be of 37 http://bmjopen.bmj.com/ 38 Chinese ethnicity, which may limit generalizability of the trial results. 39 40 41 42 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 52 BMJ Open

1 2 3 INTRODUCTION BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia 7 8 encountered in clinical practice.1 Patients with AF are at increased risk of 9 10 ischaemic stroke and systemic thromboembolism due to the formation and 11 12 embolism of left atrial thrombi, 1-3 hence long-term oral anticoagulant (OAC) for 13 14 15 thromboprophylaxis is the cornerstone in AF management. In previous 16 17 randomized clinical trials in the last century, warfarin has been shown to be 18 For peer review only 19 highly effective in reducing stroke risk compared with placebo by as much as 20 21 4 22 64% in patients with AF. More recently, non-vitamin K oral anticoagulants 23 24 (NOACs) have consistently demonstrated to be safer and more effective for 25 26 stroke prevention in patients with non-valvular AF compared with warfarin and 27 28 29 have become the recommended standard of care for the management of stroke 30 31 prevention in non-valvular AF. 32 33 34 35 While the stroke risk amongst patients with AF appears heterogeneous,5 6 36 37 http://bmjopen.bmj.com/ 38 patients with underlying valvular heart disease, particularly mitral stenosis (MS) 39 40 are at very high risk for stroke with an annual stroke risk ranging from 4% to 41 42 17% if left un-anticoagulated7 and the highest recurrences.8 However, patients 43 44 9 45 with AF and underlying MS are typically excluded in randomized control trials. on September 26, 2021 by guest. Protected copyright. 46 47 As a result, current international guidelines for management of AF do not 48 49 recommend NOACs for stroke prevention in patients with AF and underlying 50 51 3 52 moderate or severe MS. Nonetheless, off-label use of NOAC in patients with 53 54 AF and MS is not uncommon in the real-world practice. In a recently published 55 56 retrospective, observational analysis from the Republic of Korea,10 in a cohort 57 58 of 7,357 patients with MS receiving anticoagulation therapy, 35% of these 59 60

5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 52

1 2 3 patients were in fact treated with NOAC with the remaining 65% with warfarin. BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 More importantly, after propensity matching, it was shown that patients treated 7 8 with NOAC had a substantially lower risk of ischaemic stroke/systemic 9 10 embolism with an annualized risk of 2.22%/year, compared to that of 11 12 4.19%/year for patients treated with warfarin, (adjusted HR: 0.28; 95% 13 14 15 confidence interval (CI): 0.18 to 0.45), suggesting a potential role of NOAC 16 17 amongst patients with AF and underlying MS.8 10 11 18 For peer review only 19 20 21 22 This is of particular importance for Asian AF patients, in whom MS remains 23 24 relatively prevalent despite a declining trend.7 More importantly, the much 25 26 higher baseline risk of intracranial haemorrhage and apparently higher 27 28 29 ischaemic stroke risk in Asian populations potentially undermines the benefits 30 31 of warfarin therapy.12,13 14 Notably, compared with warfarin, the effectiveness, 32 33 and safety of NOACs appear to be even more superior in Asian populations 34 35 than Caucasian populations as shown in sub-analyses of pivotal randomized 36 37 http://bmjopen.bmj.com/ 38 controls trials15-17 as well as in studies using real-world data.18-23 To our 39 40 knowledge, this is the first multicentre randomized control trial aims to 41 42 comparing NOAC to warfarin to address the knowledge in stroke prevention 43 44 45 strategy in patients with AF and moderate or severe MS. This will have on September 26, 2021 by guest. Protected copyright. 46 47 immediate and long-term impacts on the management of these very high-risk 48 49 patients with AF. 50 51 52 53 54 55 56 57 58 59 60

6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 52 BMJ Open

1 2 3 METHODS AND ANALYSIS BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 This clinical trial protocol follows the Standard Protocol Items: 7 8 9 24 25 10 Recommendations for Interventional Trials (SPIRIT). The underlying 11 12 13 protocol follows the Consolidated Standards of Reporting Trials (CONSORT). 14 15 16 26 27 The study is registered with the www.ClinicalTrials.gov (NCT04045093). 17 18 For peer review only 19 20 21 22 Study Design 23 24 This is an investigator-initiated, open-label, randomized clinical trial to compare 25 26 effectiveness and safety of dabigatran 150 mg or 110 mg twice daily according 27 28 29 to kidney function with warfarin therapy with target international normalized 30 31 ratio (INR) 2-3 for stroke prevention in patients with AF and moderate or severe 32 33 MS. 34 35 36 37 http://bmjopen.bmj.com/ 38 Study Participants 39 40 Patients will be recruited from participating specialist cardiology centres in 41 42 43 Hong Kong SAR China and Mainland China. Written informed consent will be 44 45 obtained from all study participants. Table 1 summarizes the inclusion and on September 26, 2021 by guest. Protected copyright. 46 47 exclusion criteria for the study. In brief, patients with no symptoms, aged 18 48 49 years or above will be eligible if they have AF documented on standard 12-lead 50 51 52 electrocardiography (ECG) performed at screening or randomization and 53 54 moderate or severe MS as defined as the mitral valvular area (MVA) of 1.0-1.5 55 56 cm2 and <1.0 cm2, respectively. Reasons for exclusion include the presence of 57 58 59 symptoms, prosthetic valve, left atrial appendage occlusive device, and/or 60

7 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 52

1 2 3 active endocarditis; planned valvular intervention and/or planned AF ablation; BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 history of major bleeding including intracranial, intraocular, spinal or 7 8 retroperitoneal haemorrhage; unexplained anaemia with haemoglobin level 9 10 <10 g/dL, or thrombocytopenia with platelet count <100×109/L; need for 11 12 anticoagulant or antiplatelet therapy of conditions other than AF; concomitant 13 14 15 use of potent P-gp inhibitor(s) or drugs with a known interaction with dabigatran; 16 17 uncontrolled hypertension; significant kidney impairment with estimated 18 For peer review only 19 creatinine clearance (CrCl) ≤30 mL/min by the Cockcroft-Gault Formula; 28 liver 20 21 29 22 dysfunction of Child-Pugh Stage B or C; pregnancy or if there is child-baring 23 24 potential during the full duration of the study. In addition, patients considered 25 26 unsuitable by the investigator including short life expectancy <1 year due to 27 28 29 concomitant disease, substance and/or alcohol abuse or other medical 30 31 conditions. 32 33 34 35 Study Procedures 36 37 http://bmjopen.bmj.com/ 38 After providing written informed consent, all study participants will be randomly 39 40 assigned to receive dabigatran or to receive warfarin. The procedure of the trial 41 42 is summarized in figure 1, For patients randomized to receive dabigatran, the 43 44 45 dosage regimen will be determined according to the respective estimated CrCl on September 26, 2021 by guest. Protected copyright. 46 47 or if concomitantly taking interacting drugs requiring dosage adjustment. 48 49 Patients with estimated CrCl above 50 ml/min will receive dabigatran 150 mg 50 51 52 twice daily, whereas those with CrCl between 30 to 50 ml/min will receive 53 54 dabigatran 110 mg twice daily. For patients previously on warfarin randomized 55 56 to receive dabigatran, dabigatran will initiate after discontinuation of warfarin 57 58 with an INR less than or equal to 2. At the end of study, patients randomized to 59 60

8 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 52 BMJ Open

1 2 3 dabigatran will be switched back to warfarin. Warfarin will be initiated 3 days BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 prior to the termination of dabigatran. INR will be checked 5 days after initiation 7 8 of warfarin i.e., 2 days after termination of dabigatran to minimise the potential 9 10 impact of remaining dabigatran levels in elevating the INR. On the other hand, 11 12 for those randomized to receive warfarin, INR will be measured at least every 13 14 15 8 weeks with a target INR of 2.0 to 3.0. The time in therapeutic range (TTR) will 16 17 be calculated for each study participant using Rosendaal method,30 in which 18 For peer review only 19 INR will be assumed to change in a linear manner between measurements, and 20 21 22 INR values on the days without measurement are interpolated. The percentage 23 24 of time during which a study participant has an INR within 2.0-3.0 is taken as 25 26 TTR. The first follow-up visit will be scheduled 14 days after randomization and 27 28 29 then every 4 months during the study period of 1 year (Table 2). Criteria for 30 31 discontinuation or change of allocated treatment include patient request, drug 32 33 allergy, intolerable adverse drug reaction and development of other 34 35 contraindication. Patients who are randomized to receive dabigatran would be 36 37 http://bmjopen.bmj.com/ 38 switched to warfarin if CrCl is below 30 ml/min and/or develop liver dysfunction 39 40 of Child-Pugh Stage B or C. 41 42 43 44 45 Outcomes on September 26, 2021 by guest. Protected copyright. 46 47 The primary outcome is a composite of stroke or systemic embolism at 1 year. 48 49 Secondary outcomes are ischaemic stroke, systemic embolism, hemorrhagic 50 51 52 stroke, intracranial haemorrhage, major bleeding and death at 1 year. Stroke is 53 54 defined as a neurological deficit of sudden onset that persisted for more than 55 56 24 hours and corresponded to a vascular territory that cannot be explained by 57 58 other causes (such as trauma, infection or vasculitis). Stroke will be further 59 60

9 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 52

1 2 3 classified as ischaemic stroke and hemorrhagic stroke according to BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 computerized axial tomography or magnetic resonance imaging of the brain. 7 8 Intracranial haemorrhage (ICH) consists of hemorrhagic stroke (intra-cerebral 9 10 haemorrhage and cerebellar haemorrhage ), subdural haemorrhage , and 11 12 subarachnoid haemorrhage , and will be confirmed with computerized axial 13 14 15 tomography or magnetic resonance imaging of the brain. Systemic embolism 16 17 is defined as an acute vascular occlusion of an extremity or organ other than 18 For peer review only 19 the brain, documented by imaging, surgery, and/or autopsy. 20 21 22 23 24 Major bleeding is defined as a drop in the haemoglobin level of at least 2 g/dL, 25 26 transfusion of at least 2 units of blood, or symptomatic bleeding in a critical area 27 28 29 or organ. Life-threatening bleeding includes fatal bleeding, symptomatic 30 31 intracranial bleeding, bleeding with a haemoglobin drop of at least 5 g/dL, or 32 33 bleeding requiring transfusion of at least 4 units of blood or inotropic agents or 34 35 requiring surgery. All outcomes will be adjudicated by 2 independent 36 37 http://bmjopen.bmj.com/ 38 investigators in a blinded fashion. 39 40 41 42 Sample Size Calculation 43 44 45 The primary analysis is to test whether dabigatran is noninferior to warfarin for on September 26, 2021 by guest. Protected copyright. 46 47 ischaemic stroke prevention in patients with AF and moderate or severe MS. 48 49 The potential for dabigatran to preserve at least 50% of the effectiveness of 50 51 52 warfarin is considered clinically meaningful, as noninferior in patients with AF 53 54 and moderate or severe MS. The noninferiority margin is 1.49, which is derived 55 56 from the only observation study comparing vitamin K antagonist with NOAC in 57 58 patients with AF and MS.10 In the study, the annual ischaemic stroke risk of 59 60

10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 52 BMJ Open

1 2 3 patients with AF and MS receiving vitamin K antagonist and NOAC are BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 4.19%/year and 2.22%/year respectively. Accordingly, a sample size of 686 7 8 patients (343 patients in the vitamin K antagonist group and 343 in the 9 10 dabigatran group) including 10% attrition would be needed to satisfy the 11 12 noninferiority hypothesis with the upper boundary of the one-sided 95% 13 14 15 confidence interval (CI) (or equivalent with a 90% two-sided CI) and the Hazard 16 17 ratio (HR) of the primary outcome below the noninferiority margin of 1.49. 18 For peer review only 19 Hierarchical analysis for superiority will be performed if noninferiority is 20 21 22 established. 23 24 25 26 Statistical Analysis 27 28 29 Baseline data will be reported as means and standard deviations for continuous 30 31 data and as numbers and percentages for categorical data. All endpoints will 32 33 be analysed according to the intention-to-treat principle, with all patients who 34 35 undergo randomization included in the analysis. Clinical events that occur after 36 37 http://bmjopen.bmj.com/ 38 randomization and until the end of the study (at 1 year or mortality) will be 39 40 included in the primary analysis of clinical outcomes. A p-value <0.05 41 42 considered as significant. Calculations will be performed using SPSS software 43 44 45 (version 12.0). on September 26, 2021 by guest. Protected copyright. 46 47 48 49 Randomization 50 51 52 Randomization will be stratified to each study site to account for variations in 53 54 patient demographics and diagnoses. At each site, patients will be randomised 55 56 to “permuted blocks of four” (two of each study arm) to assist in equality of 57 58 numbers in each arm. An independent research officer who are blinded to this 59 60

11 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 52

1 2

3 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 study will generate the random-number table. Study staff responsible for 5 6 7 enrolment will be informed of randomization assignment by phone. Subjects 8 9 10 and clinicians will not be blinded to the randomization assignment. Data staff 11 12 13 responsible for data entry will be blinded from randomization assignment. 14 15 16 17 Data collection and management 18 For peer review only 19 20 After enrolment, each subject will be assigned a unique identifier to be used in 21 22 23 database. Data will be entered by study staff and data accuracy will be verified 24 25 26 27 by study principal investigator. Data quality control measures include queries 28 29 30 to identify missing data, outliers and discrepancies. The database will be 31 32 33 34 password protected and encrypted. Only study staff will have access to the 35 36 37 database. All paper records will be deidentified and stored securely in a locked http://bmjopen.bmj.com/ 38 39 40 41 cabinet for 5 years. Subjects who withdraw from the study will have continuous 42 43 44 monitoring stopped, usual care continued and final outcome collected for 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 analysis. 49 50 51 52 53 Data monitoring and safety 54 55 An independent Safety Committee will be established comprising of an 56 57 58 Emergency Clinician, Clinical Pharmacologist and Toxicologist. They will 59 60 receive regular reports during patient enrolment and be notified of any adverse

12 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 52 BMJ Open

1 2 3 drug reaction and study protocol violation. The Safety Committee is led by BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 Professor Bernard Cheung from the Clinical Pharmacology, Faculty of 7 8 Medicine, the University of Hong Kong, Hong Kong SAR, China. For patient 9 10 safety, discontinuation of the study is at the discretion of the cardiology clinician 11 12 to enable informed decisions to be made regarding subsequent management 13 14 15 and alternative medication use. Any medication or therapy, intervention or 16 17 procedure thought to be necessary for the safe management of the patient may 18 For peer review only 19 be administered at the discretion of the managing clinician. 20 21 22 23 24 Patient and public involvement 25 26 27 We received input from clinicians and patients which guided the design of the 28 29 30 current study and choice of research questions. No patients were directly 31 32 33 34 involved in the design of the study and choice of outcome measures. No 35 36 37 patients will be involved in recruitment or conduct of the study. Results of the http://bmjopen.bmj.com/ 38 39 40 41 study will be disseminated to subjects, the public and the scientific community. 42 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

13 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 52

1 2

3 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 ETHICS AND DISSEMINATION 5 6 7 This research protocol complies with the Declaration of Helsinki and the 8 9 International Conference on Harmonisation-Good Clinical Practice. The study 10 11 protocol has been approved by the Institutional Review Board of The University 12 13 14 of Hong Kong, and Hong Kong West Cluster, Hospital Authority, Hong Kong. 15 16 Approvals at other participating sites will be subsequently obtained before 17 18 recruitment begins.For Written peer informed review consents will only be obtained from all study 19 20 21 participants by study staff responsible for recruitment (Supplementary File 1). 22 23 Important protocol modifications will be conveyed to investigators, Institutional 24 25 26 27 Review Board, trial registries, regulators, journals and trial participants. After 28 29 30 enrolment, each subject will be assigned a unique identifier to be used in 31 32 33 34 database. Personal identity of subjects will not be used for any public purpose, 35 36 37 publication, or transmitted outside of the study team. http://bmjopen.bmj.com/ 38 39 40 41 42 43 Dataset used during the study will be available from the corresponding author 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 on reasonable request. Collaboration with other investigators will be welcomed. 48 49 50 The results of the trial will be published in peer-reviewed journals and presented 51 52 53 in conferences. 54 55 56 57 58 59 60

14 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 52 BMJ Open

1 2 3 DISCUSSION BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 To our knowledge, this is the first study to evaluate dabigatran as an alternative 7 8 to warfarin for stroke prevention in patients with AF and moderate or severe 9 10 MS. The results could fill the gap in stroke prevention strategy for this specific 11 12 group of patients with AF with immediate and long-term impacts on clinical 13 14 15 practice. 16 17 18 For peer review only 19 AF is the most commonly encountered sustained cardiac arrhythmia in clinical 20 21 1 31 22 practice. While patients with AF have in general increased risk of stroke, 4 23 24 subgroups are at particularly high risk necessitating long-term anticoagulation 25 26 therapy. These include (1) patients with non-valvular AF and high CHA2DS2- 27 28 29 VASc score, (2) patients with AF and hypertrophic cardiomyopathy, (3) patients 30 31 with AF and MS, and (4) patients with AF and mechanical heart valves. (Figure 32 33 2) In the past decade, NOAC has emerged as the preferred agent over warfarin 34 35 for stroke prevention in patients with non-valvular AF and high CHA DS -VASc 36 2 2 37 http://bmjopen.bmj.com/ 38 score. In the 4 pivotal studies comparing NOACs and warfarin in patients with 39 40 non-valvular AF, NOACs are at least as effective as warfarin to reduce stroke 41 42 and at the same time are much safer alternatives in terms of life-threatening 43 44 45 bleeding complications. Nonetheless, direct extrapolation of these results to on September 26, 2021 by guest. Protected copyright. 46 47 other subgroups of AF patients may not be appropriate, given the different 48 49 mechanisms of thrombus formation in different diseases. For instance, in the 50 51 52 RE-ALIGN trial randomizing 252 patients with recent mechanical valvular 53 54 replacement in a 2:1 ratio to receive either dabigatran or warfarin, there was an 55 56 excess of thromboembolic as well as bleeding events among patients 57 58 randomized to dabigatran, rendering the study prematurely terminated.32 59 60

15 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 52

1 2

3 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 On the other hand, international guidelines do not recommend NOACs for 7 8 patients with AF and moderate or severe mitral stenosis due to the lack of 9 10 reliable data from clinical trials. Nonetheless, it remains undetermined whether 11 12 NOACs can be used as an alternative to warfarin for patients with AF and 13 14 15 moderate or severe MS due to the lack of clinical trial. The current study has 16 17 several important implications, particularly in Asian countries. First, while MS is 18 For peer review only 19 now a rare condition in developed countries, it remains relatively prevalent in 20 21 22 many Asian countries. In addition, the risk of stroke amongst patients with AF 23 24 and MS is only second to those with mechanical valvular replacement ranging 25 26 from 4 to 17%. Second, previous epidemiological studies12,13 14 21 33 34 and sub- 27 28 15-17 29 analyses of the pivotal NOAC trials have consistently reported a much 30 31 higher nominal risk of ICH amongst Asians than non-Asians, favouring NOACs 32 33 over warfarin therapy. More importantly, the notoriously poor time in therapeutic 34 35 range (TTR) for warfarin in Asian populations observed in real-world data18 34- 36 37 http://bmjopen.bmj.com/ 38 38 and pivotal NOAC trials15-17 substantially undermines the overall clinical 39 40 benefits of warfarin therapy. In fact, the annual incidence of ICH amongst 41 42 patients with AF and MS treated with warfarin has been reported to be as high 43 44 10 45 as 0.93% per year, urging a much safer alternative. on September 26, 2021 by guest. Protected copyright. 46 47 48 49 In the present study, the NOAC of choice is dabigatran, the first NOAC with an 50 51 52 approved indication for stroke prevention in non-valvular AF patients from the 53 54 United States Food and Drug Administration in 2009. In the pivotal study, the 55 56 RE-LY study,39 patients with non-valvular AF with CHADS2>1 were randomly 57 58 assigned to two doses of dabigatran: 110 mg or 150 mg twice daily; or adjusted- 59 60

16 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 52 BMJ Open

1 2 3 dose warfarin. After a median follow-up of 2.0 years, the low-dose regime was BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 found to be as effective as warfarin in preventing the primary endpoint (a 7 8 composite of stroke and systemic embolism) (1.52%/year vs. 1.69%/year), but 9 10 with a substantially lower risk of major bleeding and ICH.39 On the other hand, 11 12 the standard dose dabigatran (150mg BD) is superior to warfarin in reducing 13 14 15 the primary composite endpoint and ICH, with a comparable risk of major 16 17 bleeding.39 In an analysis comparing the effectiveness and safety of dabigatran 18 For peer review only 19 according to the ethnicity of study participants, dabigatran appears to be more 20 21 22 effective in stroke prevention as well as safer in terms of ICH compared with 23 24 warfarin. This is in concordance to subsequent real-world cohorts of AF patients 25 26 from territory-wide registries from Asia Pacific region. Plausible explanations 27 28 29 include suboptimal quality of warfarin therapy with low time in therapeutic range 30 31 and a higher risk of ICH in Asian populations.18 23 35 40 41 An additional reason 32 33 for the choice of dabigatran in the present study is the wide availability of its 34 35 antidote, idarucizumab in Asian countries, which provides extra-protection of 36 37 http://bmjopen.bmj.com/ 38 patients in the clinical trial. 39 40 41 42 The study is designed to provide clinicians with robust, much-needed 43 44 45 information regarding stroke prevention strategy for patients with AF and on September 26, 2021 by guest. Protected copyright. 46 47 moderate or severe MS. The results will have immediate and long-term impacts 48 49 on the management of these very high-risk patients with AF. 50 51 52 53 54 55 56 57 58 59 60

17 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 52

1 2 3 Author contributions: BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 MZ, EWC, JH, CKW, BPY and CWS contributed to the conception and design 7 8 of the study. MZ, EWC, JH, CKW, YML, DH, CCL, CCT, AYTW, ASYY, KKWC, 9 10 YF, NT, JYC, CYY, KLL, CWC, HL, AN, KF, MHJ, KHY, BPY and CWS 11 12 13 contributed to the acquisition of data. Data analysis and interpretation will be 14 15 16 conducted by MZ, EWC, JH, CKW, BPY and CWS. MZ, EWC, JH, CKW, BPY 17 18 For peer review only 19 20 and CWS wrote first draft of the protocol and revised the protocol critically for 21 22 23 important intellectual content. All authors have read and approved the final 24 25 26 27 version of the manuscript to be published. 28 29 30 31 32 Acknowledgement: 33 34 35 None. 36 37 http://bmjopen.bmj.com/ 38 39 Funding statement: 40 41 None. 42 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 Conflict of interest: 47 48 None. 49 50 51 52 53 54 55 56 57 58 59 60

18 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 52 BMJ Open

1 2 3 Legends: BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 Figure 1. Design of the DAVID-MS study. 7 8 Figure 2. Four main groups of patients with AF requiring long-term 9 10 anticoagulation therapy. MS: mitral stenosis; NOAC: non-vitamin K oral 11 12 anticoagulant; and VKA: vitamin K antagonist. 13 14 15 16 Supplementary Files: 17 18 For peer review only 19 20 Supplementary File 1: Model consent form in English 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

19 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 52

1 2 3 Tables: BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 5 6 Table 1. Inclusion and Exclusion Criteria 7 8 Inclusion Criteria 9 10  Patients with AF documented with standard 12-lead ECG documented AF 11 12 on the day of screening or randomization 13  Patients with age >18 years 14 15  Patients with moderate or severe MS i.e., MVA <1.5 cm2 16  Patients should be able to provide a written, informed consent. 17 18  PatientsFor should peerhave all 4 inclusion-criteria review fulfilled only to be qualified for the 19 20 study. 21 Exclusion criteria 22 23  Patients with prosthetic valve, or with active endocarditis 24 25  Patients with heart failure symptom 26  Patients with planned valvular intervention within 1 year 27 28  Patients with left atrial appendage occlusive device 29 30  Patients with planned AF ablation 31  Patients with a history of intracranial, intraocular, spinal, or retroperitoneal 32 33 bleeding 34 35  Unexplained anemia (haemoglobin level <10 g/dL) or thrombocytopenia 36 (platelet count <100×109/L) 37 http://bmjopen.bmj.com/ 38  Need for anticoagulant therapy of disorders other than AF 39  Patients receiving antiplatelet therapy for disorders other than AF 40 41  Patients receiving concomitant P-gp inhibitors and/or medications known to 42 43 interact with dabigatran 44  Uncontrolled hypertension (systolic blood pressure >180 mmHg and/or 45 on September 26, 2021 by guest. Protected copyright. 46 diastolic blood pressure >100 mmHg) 47 48  Estimated creatinine clearance ≤30 mL/min 49  Liver dysfunction of Child-Pugh stage B or C 50 51  Women who are pregnant or of childbearing potential who refuse to use a 52 medically acceptable form of contraception throughout the study 53 54  Patients considered unreliable by the investigator or have a life expectancy 55 56 less than 1 year because of concomitant disease, or has any condition, 57 which in the opinion of the investigator, would not allow safe participation in 58 59 the study (e.g. drug addiction, alcohol abuse). 60

20 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 52 BMJ Open

1 2 3 Abbreviations: AF: atrial fibrillation; MS: mitral stenosis; MVA: mitral valvular BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 area 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

21 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open Page 24 of 52

1 2 3 4 5 Table 2. Study visits 6 Visits -1 0 1 2 3 4 5 6 7 UNS** EOS*** 7 8 Weeks -2 to 0 0 8 16 24 32 40 48 56 -- 56-60 9 10 Informed consent X 11 12 Inclusion & exclusion criteriaFor peerX review only 13 14 15 Randomization X 16 17 Medical history X http://bmjopen.bmj.com/ 18 19 Physical examination X X X X X X X X X X 20 21 22 Echocardiography X 23 24 INR X X* X* X* X* X* X* X* X* X 25 on September 26, 2021 by guest. Protected copyright. 26 Renal function X X X X X X 27 28 29 Drug dispensing X X X X X X X X 30 31 Drug collection X X X X X X X X 32 33 Outcome events X X X X X X X X X 34 35 36 Adverse events X X X X X X X X X 37 38 39 40 41 42 22 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Page 25 of 52 BMJ Open

17 http://bmjopen.bmj.com/ 18 19 20 21 22 23 24

25 on September 26, 2021 by guest. Protected copyright. 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 23 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 26 of 52

1 2

3 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 REFERENCES 5 6 7 1. Lip GY, Tse HF, Lane DA. Atrial fibrillation. Lancet 2012;379(9816):648-61. 8 doi: S0140-6736(11)61514-6 [pii] 9 10.1016/S0140-6736(11)61514-6 [published Online First: 2011/12/15] 10 2. January CT, Wann LS, Calkins H, et al. 2019 AHA/ACC/HRS Focused 11 Update of the 2014 AHA/ACC/HRS Guideline for the Management of 12 Patients With Atrial Fibrillation: A Report of the American College of 13 14 Cardiology/American Heart Association Task Force on Clinical Practice 15 Guidelines and the Heart Rhythm Society in Collaboration With the 16 Society of Thoracic Surgeons. Circulation 2019;140(2):e125-e51. doi: 17 10.1161/CIR.0000000000000665 [published Online First: 2019/01/29] 18 3. Kirchhof P, ForBenussi peerS, Kotecha review D, et al. 2016 ESConly Guidelines for the 19 management of atrial fibrillation developed in collaboration with 20 21 EACTS. Europace 2016;18(11):1609-78. doi: 22 10.1093/europace/euw295 [published Online First: 2016/11/04] 23 4. Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotic therapy to 24 prevent stroke in patients who have nonvalvular atrial fibrillation. Ann 25 Intern Med 2007;146(12):857-67. doi: 10.7326/0003-4819-146-12- 26 200706190-00007 [published Online First: 2007/06/20] 27 5. Siu CW, Pong V, Zhang X, et al. Risk of ischemic stroke after new-onset 28 29 atrial fibrillation in patients with hyperthyroidism. Heart Rhythm 30 2009;6(2):169-73. doi: 10.1016/j.hrthm.2008.10.023 [published Online 31 First: 2009/02/04] 32 6. Guo Y, Wang H, Tian Y, et al. Multiple risk factors and ischaemic stroke in 33 the elderly Asian population with and without atrial fibrillation. An 34 analysis of 425,600 Chinese individuals without prior stroke. Thromb 35 Haemost 2016;115(1):184-92. doi: 10.1160/th15-07-0577 [published 36 37 Online First: 2015/09/01] http://bmjopen.bmj.com/ 38 7. Chandrashekhar Y, Westaby S, Narula J. Mitral stenosis. Lancet 39 2009;374(9697):1271-83. doi: 10.1016/S0140-6736(09)60994-6 40 8. De Caterina R, John Camm A. Non-vitamin K antagonist oral 41 anticoagulants in atrial fibrillation accompanying mitral stenosis: the 42 concept for a trial. Europace 2016;18(1):6-11. doi: 43 44 10.1093/europace/euv288 [published Online First: 2015/10/10] 45 9. Breithardt G, Baumgartner H, Berkowitz SD, et al. Clinical characteristics on September 26, 2021 by guest. Protected copyright. 46 and outcomes with rivaroxaban vs. warfarin in patients with non- 47 valvular atrial fibrillation but underlying native mitral and aortic valve 48 disease participating in the ROCKET AF trial. Eur Heart J 49 2014;35(47):3377-85. doi: 10.1093/eurheartj/ehu305 [published Online 50 First: 2014/08/26] 51 52 10. Kim JY, Kim SH, Myong JP, et al. Outcomes of Direct Oral Anticoagulants 53 in Patients With Mitral Stenosis. J Am Coll Cardiol 2019;73(10):1123- 54 31. doi: 10.1016/j.jacc.2018.12.047 55 11. Giugliano RP, O'Gara PT. DOACs in Patients With Mitral Stenosis and 56 Atrial Fibrillation: Time for a Randomized Clinical Trial. J Am Coll 57 Cardiol 2019;73(10):1132-34. doi: 10.1016/j.jacc.2018.12.048 58 59 60

24 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 52 BMJ Open

1 2 3 12. Tse HF, Wang YJ, Ai-Abdullah MA, et al. Stroke Prevention in Atrial BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 Fibrillation - An Asian Stroke Perspective. Heart Rhythm 2013 doi: 5 6 10.1016/j.hrthm.2013.03.017 [published Online First: 2013/03/19] 7 13. Chong BH, Chan KH, Pong V, et al. Use of aspirin in Chinese after 8 recovery from primary intracranial haemorrhage. Thromb Haemost 9 2012;107(2):241-7. doi: 10.1160/th11-06-0439 [published Online First: 10 2011/12/22] 11 14. Friberg L, Rosenqvist M, Lip GY. Evaluation of risk stratification schemes 12 for ischaemic stroke and bleeding in 182 678 patients with atrial 13 14 fibrillation: the Swedish Atrial Fibrillation cohort study. Eur Heart J 15 2012;33(12):1500-10. doi: ehr488 [pii] 16 10.1093/eurheartj/ehr488 [published Online First: 2012/01/17] 17 15. Hori M, Connolly SJ, Zhu J, et al. Efficacy and safety of dabigatran versus 18 warfarinFor in patients peer with atrial review fibrillation: analysis only in Asian population in 19 RE-LY trial. Cerebrovascular Disease 2012;34 (Suppl 1)(Asia Pacific 20 21 Stroke Conference 2012):9. 22 16. Hankey GJ, Stevens S, Piccini JP, et al. Predictors of Intracranial 23 Hemorrhage among anticoagulated patients with atrial fibrillation: 24 insights from the Rivaroxaban once daily oral direct factor Xa inhibition 25 compared 26 with vitamin K antagonism for prevention of stroke and embolism trial in atrial 27 fibrillation (ROCKET AF). Stroke 2012;43:A152. 28 29 17. Tanahashi N, Hori M, Matsumoto M, et al. Rivaroxaban versus Warfarin in 30 Japanese Patients with Nonvalvular Atrial Fibrillation for the Secondary 31 Prevention of Stroke: A Subgroup Analysis of J-ROCKET AF. J Stroke 32 Cerebrovasc Dis 2013 doi: S1052-3057(12)00437-5 [pii] 33 10.1016/j.jstrokecerebrovasdis.2012.12.010 [published Online First: 34 2013/01/29] 35 18. Chan PH, Huang D, Lau CP, et al. Net Clinical Benefit of Dabigatran Over 36 37 Warfarin in Patients With Atrial Fibrillation Stratified by CHA2DS2- http://bmjopen.bmj.com/ 38 VASc and Time in Therapeutic Range. Can J Cardiol 2016;32(10):1247 39 e15-47 e21. doi: 10.1016/j.cjca.2016.01.016 40 19. Chan PH, Li WH, Hai JJ, et al. Gastrointestinal haemorrhage in atrial 41 fibrillation patients: impact of quality of anticoagulation control. Eur 42 Heart J Cardiovasc Pharmacother 2015;1(4):265-72. doi: 43 44 10.1093/ehjcvp/pvv032 [published Online First: 2016/08/18] 45 20. Lee YK, Lau YM, Cai ZJ, et al. Modeling Treatment Response for Lamin on September 26, 2021 by guest. Protected copyright. 46 A/C Related Dilated Cardiomyopathy in Human Induced Pluripotent 47 Stem Cells. J Am Heart Assoc 2017;6(8) doi: 10.1161/jaha.117.005677 48 [published Online First: 2017/07/30] 49 21. Lau WCY, Li X, Wong ICK, et al. Bleeding-related hospital admissions and 50 30-day readmissions in patients with non-valvular atrial fibrillation 51 52 treated with dabigatran versus warfarin. J Thromb Haemost 53 2017;15(10):1923-33. doi: 10.1111/jth.13780 54 22. Huang D, Cheng YY, Chan PH, et al. Rationale and design of the 55 screening of pulmonary hypertension in systemic lupus erythematosus 56 (SOPHIE) study. ERJ Open Res 2018;4(1) doi: 57 10.1183/23120541.00135-2017 [published Online First: 2018/03/14] 58 23. Qi X, Wong BL, Lau SH, et al. A hemoglobin-based oxygen carrier 59 60 sensitized Cisplatin based chemotherapy in hepatocellular carcinoma.

25 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 52

1 2 3 Oncotarget 2017;8(49):85311-25. doi: 10.18632/oncotarget.19672 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 [published Online First: 2017/11/22] 5 6 24. Chan AW, Tetzlaff JM, Altman DG, et al. SPIRIT 2013 statement: defining 7 standard protocol items for clinical trials. Ann Intern Med 8 2013;158(3):200-7. doi: 10.7326/0003-4819-158-3-201302050-00583 9 [published Online First: 2013/01/09] 10 25. Chan AW, Tetzlaff JM, Gotzsche PC, et al. SPIRIT 2013 explanation and 11 elaboration: guidance for protocols of clinical trials. BMJ 12 2013;346:e7586. doi: 10.1136/bmj.e7586 [published Online First: 13 14 2013/01/11] 15 26. Altman DG, Schulz KF, Moher D, et al. The revised CONSORT statement 16 for reporting randomized trials: explanation and elaboration. Ann Intern 17 Med 2001;134(8):663-94. doi: 10.7326/0003-4819-134-8-200104170- 18 00012 [publishedFor peer Online First: review 2001/04/17] only 19 27. Moher D, Schulz KF, Altman D, et al. The CONSORT Statement: revised 20 21 recommendations for improving the quality of reports of parallel-group 22 randomized trials 2001. Explore (NY) 2005;1(1):40-5. doi: 23 10.1016/j.explore.2004.11.001 [published Online First: 2006/06/24] 24 28. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum 25 creatinine. Nephron 1976;16(1):31-41. doi: 10.1159/000180580 26 [published Online First: 1976/01/01] 27 29. Pugh RN, Murray-Lyon IM, Dawson JL, et al. Transection of the 28 29 oesophagus for bleeding oesophageal varices. Br J Surg 30 1973;60(8):646-9. doi: 10.1002/bjs.1800600817 [published Online 31 First: 1973/08/01] 32 30. Rosendaal FR, Cannegieter SC, van der Meer FJ, et al. A method to 33 determine the optimal intensity of oral anticoagulant therapy. Thromb 34 Haemost 1993;69(3):236-9. 35 31. Camm AJ, Kirchhof P, Lip GY, et al. Guidelines for the management of 36 37 atrial fibrillation: the Task Force for the Management of Atrial http://bmjopen.bmj.com/ 38 Fibrillation of the European Society of Cardiology (ESC). Europace 39 2010;12(10):1360-420. doi: euq350 [pii] 40 10.1093/europace/euq350 [published Online First: 2010/09/30] 41 32. Eikelboom JW, Connolly SJ, Brueckmann M, et al. Dabigatran versus 42 warfarin in patients with mechanical heart valves. N Engl J Med 43 44 2013;369(13):1206-14. doi: 10.1056/NEJMoa1300615 [published 45 Online First: 2013/09/03] on September 26, 2021 by guest. Protected copyright. 46 33. Chan EW, Lau WC, Siu CW, et al. Effect of suboptimal anticoagulation 47 treatment with antiplatelet therapy and warfarin on clinical outcomes in 48 patients with nonvalvular atrial fibrillation: A population-wide cohort 49 study. Heart Rhythm 2016;13(8):1581-8. doi: 50 10.1016/j.hrthm.2016.03.049 51 52 34. Teo KC, Mahboobani NR, Lee R, et al. Warfarin associated intracerebral 53 hemorrhage in Hong Kong Chinese. Neurol Res 2014;36(2):143-9. doi: 54 10.1179/1743132813Y.0000000275 55 35. Huang D, Wong CL, Cheng KW, et al. Impact of provision of time in 56 therapeutic range value on anticoagulation management in atrial 57 fibrillation patients on warfarin. Postgrad Med J 2018;94(1110):207-11. 58 doi: 10.1136/postgradmedj-2017-135457 59 60

26 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 52 BMJ Open

1 2 3 36. Chan PH, Hai JJ, Chan EW, et al. Use of the SAMe-TT2R2 Score to BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 Predict Good Anticoagulation Control with Warfarin in Chinese Patients 5 6 with Atrial Fibrillation: Relationship to Ischemic Stroke Incidence. PLoS 7 One 2016;11(3):e0150674. doi: 10.1371/journal.pone.0150674 8 37. Ho CW, Ho MH, Chan PH, et al. Ischemic stroke and intracranial 9 hemorrhage with aspirin, dabigatran, and warfarin: impact of quality of 10 anticoagulation control. Stroke 2015;46(1):23-30. doi: 11 10.1161/STROKEAHA.114.006476 12 38. Siu CW, Tse HF. Net clinical benefit of warfarin therapy in elderly Chinese 13 14 patients with atrial fibrillation. Circ Arrhythm Electrophysiol 15 2014;7(2):300-6. doi: 10.1161/CIRCEP.113.000858 16 39. Connolly SJ, Ezekowitz MD, Yusuf S, et al. Dabigatran versus warfarin in 17 patients with atrial fibrillation. N Engl J Med 2009;361(12):1139-51. doi: 18 10.1056/NEJMoa0905561For peer review only 19 40. Hai JJ, Chan PH, Chan YH, et al. Prediction of Thromboembolic Events in 20 21 Heart Failure Patients in Sinus Rhythm: The Hong Kong Heart Failure 22 Registry. PLoS One 2016;11(12):e0169095. doi: 23 10.1371/journal.pone.0169095 [published Online First: 2016/12/31] 24 41. Chan PH, Li WH, Hai JJ, et al. Time in Therapeutic Range and 25 Percentage of International Normalized Ratio in the Therapeutic Range 26 as a Measure of Quality of Anticoagulation Control in Patients With 27 Atrial Fibrillation. Can J Cardiol 2016;32(10):1247 e23-47 e28. doi: 28 29 10.1016/j.cjca.2015.10.029 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

27 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 52 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Figure 1. Design for the DAVID-MS study

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 52 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

1 2 3 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 Information sheet 5 6 Study Name: Rationale and design of Dabigatran for Mitral Stenosis Atrial Fibrillation 7 8 Trial 9 Version no.: v.1.2 (18/Nov/2019) 10 11 12 Protocol no.: DAMS-01 13 14 Protocol version no.: v.1.2 (18/Nov/2019) 15 Study site: Queen Mary Hospital 16 17 Study Principal Investigator: Prof. SIU Chung Wah David 18 ______For peer review only ______19 20 You are being invited to take part in a research study. Before you decide, it is important to 21 understand why the research is being done and what it will involve. Please take time to read the 22 23 following information carefully. Ask the study doctor or research staff any questions you may 24 have before signing the attached consent form. 25 26 27 About This Study 28 29 The purpose of our study is to find out the efficacy and safety of non-vitamin K oral 30 anticoagulants (NOACs), the drugs used to prevent ischemic stroke for atrial fibrillation (AF) 31 32 patients. 33 34 35 Why have I been chosen? 36 You are suffering from atrial fibrillation, a heart disease associated with 5-fold increase in 37 http://bmjopen.bmj.com/ 38 ischemic stroke risk. Currently, NOAC is one of the most effective drug groups in preventing 39 ischemic stroke under this condition. Among these AF patients, some have underlying valvular 40 41 heart diseases with particularly high risk for stroke for those with mitral stenosis (MS) and the 42 annual stroke rate ranges from 4% to 17% if left un-anticoagulated. 43 44 45 However, there’s lack of research-based evidence to indicate the efficacy and safety of NOACs on September 26, 2021 by guest. Protected copyright. 46 47 for patients with both AF and MS. As a result, there’re still no standard guidelines for stroke 48 prevention management regarding to NOACs for AF patients with underlying moderate to severe 49 50 MS. On the other hand, there’s recent foreign study suggesting the potential role of NOACs 51 amongst AF patients with underlying MS in stroke prevention. 52 53 54 Concerning the very high risk for stroke for AF patients with underlying MS, also higher 55 56 baseline risk of intracranial haemorrhage and higher ischemic stroke risk in Asian populations, 57 we launch this study aiming at comparing the efficacy and safety of one of the NOACs – 58 59 Dabigatran (150mg or 110mg according to subjects’ renal function) – with normal warfarin 60 [1]

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 52 BMJ Open

1 2 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 3 therapy in AF patients with moderate or severe MS. We plan to recruit a total of 686 subjects 4 5 randomizing into 2 groups of investigational Dabigatran and warfarin in a 1:1 ratio. Since you 6 have carried both heart problems, you are invited into our study. 7 8 9 What will happen to me if I take part? 10 11 If you meet the criteria of this study and are being enrolled, our investigator(s) shall have a short 12 interview with you (less than 10 minutes) to explain the benefits and potential side effects of 13 14 NOACs. Enough time will be given for understanding and solving any queries raised, and 15 written consent has to be signed for agreement of study participation. You will then be 16 17 randomized into either Dabigatran or warfarin group, which is open to your notice. 18 For peer review only 19 20 Study period of individual participants will be around 1 year. We will obtain medical history 21 directly from you, hospital record as well as electronic medical record under Hospital Authority. 22 23 Within the study time frame, you will have the first follow-up at 2-week interval after 24 randomization. After that, we will arrange regular follow-ups of every 4 months for you to 25 26 monitor the effect and safety of the drug prescribed until the study ends. We will perform certain 27 investigations during study visits, including physical examination, echocardiography (during the 28 29 first visit only) and blood sampling via venipuncture. You will be responsible to comply with the 30 scheduled study visits, study procedures and prescription plan, and report to us as soon as 31 32 possible for any adverse effects appear. 33 34 35 There are no extra expenses anticipated for participating in the clinical trial. You simply need to 36 pay for the regular specialty follow-up fee and the regular medication fee under Hospital 37 http://bmjopen.bmj.com/ 38 Authority policy as usual for each time scheduled or unscheduled follow-ups. On the other hand, 39 there will not have reimbursement in any forms from the study. 40 41 42 What are the benefits of participating? 43 44 NOAC is currently a self-financing item under Hospital Authority. That means patients need to 45 purchase the drug themselves or only patients meet certain medical criteria will the item be free. on September 26, 2021 by guest. Protected copyright. 46 47 In this study, according to randomization, you will be given free-of-charged NOAC for stroke 48 prevention secondary to AF, which is significantly safe and efficacious over the traditional 49 50 warfarin therapy. Close monitoring by experienced medical staff will be held to ensure your 51 safety. 52 53 54 Besides, your contribution is important to provide valuable information for stroke prevention 55 56 strategy for patients with mitral stenosis and that may be immediately translatable to real clinical 57 practice. It may also provide necessary evidence for establishing relevant universal guidelines. 58 59 60 [2]

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 52

1 2 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 3 What if something goes wrong? 4 5 Both dabigatran and warfarin are registered medications under the Pharmacy and Poisons 6 Ordinance (Cap. 138) in Hong Kong Special Administrative Region. They have been overseen 7 8 for their safety, efficacy and quality. Being randomized into either group (a 50/50 chance like 9 flipping a coin) in this study, you will be prescribed corresponding anticoagulant with dosage 10 11 adjustment based on your coagulation or renal blood-check result, according to standard medical 12 guidelines. 13 14 15 As with all other researches regarding to clinical trial, there may involve harms and risks that are 16 17 already known or currently unknown and unforeseen with the drug treatment. You are free to 18 raise queries and concernsFor topeer our investigators review prior to consentingonly and at any time during the 19 20 study. Our medical staff will closely monitor your condition throughout the whole study period 21 and you are responsible to tell our research staff as soon as possible for any changes in medical 22 23 condition. Below are listed known side effects of the two anticoagulants. 24 25 26 Side effects of Dabigatran: 27 Common – nausea / diarrhea / indigestion / stomach upset / stomach pain / stomach burn / 28 29 unexpected bruising / minor bleeding 30 Less common or rare – allergy / skin rash / itchiness / headache / dizziness / weakness / 31 32 unexpected or uncontrollable bleeding / coffee-ground vomiting / 33 brown urine / black stool / swelling / pain 34 35 36 Side effects of Warfarin: 37 http://bmjopen.bmj.com/ 38 Common – unexpected bruising / minor bleeding / bloating / nausea / vomiting / diarrhea / 39 loss of appetite 40 41 Less common or rare – allergy / skin rash / itchiness / headache / dizziness / weakness / 42 unexpected or uncontrollable bleeding / coffee-ground vomiting / 43 44 brown urine / black stool / swelling / pain 45 on September 26, 2021 by guest. Protected copyright. 46 47 Facts between Dabigatran and Warfarin: 48 49 Dabigatran Warfarin 50 No need for regular blood-check Regular blood-check 51 52 Fixed dosage Regular dosage adjustment based on blood 53 result 54 55 Not much food avoidance A number of food that can affect drug 56 efficacy has to be avoided 57 58 59 We indeed do not expect significant harms related to your participation to the study. In the 60 [3]

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 52 BMJ Open

1 2 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 3 unlikely event of harm resulting directly from your participation in this study, medical treatment 4 5 will be provided. Discontinuation of study treatment depends on discretion of investigator based 6 on your medical condition, subsequent management and alternative medication use. Your 7 8 willingness will be taken into consideration and prioritize. We are open to discussion to your 9 concern and you definitely have the rights at any time to informedly withdraw from the study. 10 11 There are no special compensation arrangements provided to you in this study. If you are harmed 12 due to someone’s negligence, you may have grounds for a legal action but you may have to pay 13 14 for it. Regardless of this, if you wish to complain about any aspects of the way you have been 15 approached or treated during the course of this study, the normal health service complaint 16 17 mechanisms will be available to you. 18 For peer review only 19 20 If you have any queries related to the insurance coverage from your own insurer(s) for your 21 participation in the study, please discuss with your insurance consultant(s). 22 23 24 What are the alternatives for treatment? 25 26 Your participation in this study is absolutely voluntary. You may choose not to participate in this 27 study by simply telling our research staff. If you decline this study, your medical appointments 28 29 and medications will remain unchanged, or you may have to take alternative medical advice 30 from doctor(s). You also have the rights at any time to withdraw from the study. In this case, we 31 32 may arrange a final study visit for assessing and monitoring your health status. Your future 33 follow-up appointments will be scheduled and conducted as directed by your physician. Your 34 35 decision will not in any way affect your medical care or treatments. 36 37 http://bmjopen.bmj.com/ 38 What if new information becomes available? 39 During the course of the study, if any new information becomes available that may affect 40 41 investigators’ medical decision and/or relate to your willingness to continue to participate in this 42 study, your research doctor will tell you about it in a timely manner and discuss with you. You 43 44 would have the rights of access to personal data and known study results, if and when needed. 45 on September 26, 2021 by guest. Protected copyright. 46 47 There are no foreseeable circumstances that the study will be ended unintentionally. Unless there 48 is safety concern of the investigational drug from relative studies or from drug manufacturers, 49 50 the study will be held according to protocol. In case of official mid-way termination of study, 51 participants will be arranged similarly as of study discontinuation, with additional medical 52 53 assessment and treatments as required to ensure patient safety. 54 55 56 Will my participation in this study be kept confidential? 57 As a subject in this research study, all your information will be kept confidential. Your name or 58 59 your personal identity will not be used for any public purposes, publications, or transmitted 60 [4]

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 52

1 2 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 3 outside of the medical centre. Under the laws of the Hong Kong Special Administrative Region 4 5 and, in particular, the Personal Data (Privacy) Ordinance (Cap. 486), you enjoy or may enjoy 6 rights for the protection of the confidentiality of your personal data, such as those regarding to 7 8 the collection, custody, retention, management, control, use (including analysis or comparison), 9 transfer in or out of Hong Kong, non-disclosure, erasure and/or in any way dealing with or 10 11 disposing of any of your personal data in or for this study. 12 13 14 By consenting to participate in this study, you expressly authorize the access to, the use of, and 15 the retention of your personal data by the investigator(s) and members of his research team, 16 17 representatives of the sponsor, and Institutional Review Board of the University of Hong Kong / 18 Hospital Authority ForHong Kong peer West Cluster review (HKU/HA HKWonly IRB) for the purposes and in the 19 20 manner described in this informed consent process. 21 22 23 By consenting to participate in this study, you also expressly authorize relevant government 24 agencies (e.g. Hong Kong Department of Health) to get access to your personal data for the 25 26 purpose of checking and verifying the integrity of study data and assessing compliance with the 27 study protocol and other relevant requirements. 28 29 30 For any queries, you should consult the Privacy Commissioner for Personal Data or his office 31 32 (tel no.: 852-2827-2827) as to the proper monitoring or supervision of your personal data 33 protection so that your full awareness and understanding of the significance of compliance with 34 35 the law governing privacy data is assured. 36 37 http://bmjopen.bmj.com/ 38 Who should I contact if have questions? 39 If you have any questions regarding to this study, you may contact Dr. Siu Chung Wah at 40 41 852-2255-3597. If you have any queries regarding to your rights in the study, you may contact 42 the Secretary of HKU/HA HKU IRB at 852-2255-4086. 43 44 45 on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 [5]

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 52 BMJ Open

1 2 3 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from 4 Consent Form 5 6 Study Name: Rationale and design of Dabigatran for Mitral Stenosis Atrial Fibrillation 7 8 Trial 9 Study Principal Investigator: Prof. SIU Chung Wah David 10 11 12 By signing below, I agree that: 13 14 1. I confirm that I have read and understood the information sheet for the above study and have 15 had the opportunity to ask questions. 16 17 2. I understand that my participation is voluntary and that I am free to withdraw at any time, 18 without giving anyFor reasons, peer without my review medical care or onlylegal rights being affected. 19 20 3. I understand that sections of any of my medical notes may be looked at by responsible 21 individuals or from regulatory authorities where it is relevant to my taking part in research. I 22 23 give permission for these individuals to have access to my records. 24 4. I agree to take part in the above study. 25 26 27 28 29 30 31 32 33 34 35 ______36 Participant’s signature Participant’s name Date 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 ______45 Witness’s signature Witness’s name Date on September 26, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 ______54 Investigator’s signature Investigator’s name Date 55 56 57 58 59 60 [6]

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open Page 38 of 52

1 2 3 SPIRIT 2013 Checklist: Recommended items to address in a clinical trial protocol and related documents 4 5 6 7 Section/item ItemNo Description Page 8 9 Administrative information 10 11 Title 1 Descriptive title identifying the study design, 1 12 For peerpopulation, review interventions, and, onlyif applicable, trial 13 acronym 14 15 Trial registration 2a Trial identifier and registry name. If not yet 3, 7 16 registered, name of intended registry 17 http://bmjopen.bmj.com/ 18 2b All items from the World Health Organization Uploaded to BMJ Open 19 20 Trial Registration Data Set server 21 22 Protocol version 3 Date and version identifier 1 23 24 Funding 4 Sources and types of financial, material, and 18

25 other support on September 26, 2021 by guest. Protected copyright. 26 27 Roles and responsibilities 5a Names, affiliations, and roles of protocol 1, 15 28 contributors 29 30 5b Name and contact information for the trial 18 31 sponsor 32 33 34 35 36 37 38 39 40 41 42 1 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Page 39 of 52 BMJ Open

1 2 3 4 5c Role of study sponsor and funders, if any, in 18 5 study design; collection, management, analysis, 6 and interpretation of data; writing of the report; 7 8 and the decision to submit the report for 9 publication, including whether they will have 10 ultimate authority over any of these activities 11 12 For5d peerComposition, review roles, and responsibilities only of the 12 13 coordinating centre, steering committee, 14 15 endpoint adjudication committee, data 16 management team, and other individuals or

17 groups overseeing the trial, if applicable (see http://bmjopen.bmj.com/ 18 Item 21a for data monitoring committee) 19 20 21 22 23 24

25 on September 26, 2021 by guest. Protected copyright. 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 2 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open Page 40 of 52

1 2 3 4 5 Introduction 6 7 Background and rationale 6a Description of research question and justification 5-6 8 9 for undertaking the trial, including summary of 10 relevant studies (published and unpublished) 11 examining benefits and harms for each 12 For peerintervention review only 13 14 6b Explanation for choice of comparators 5-6 15 16 Objectives 7 Specific objectives or hypotheses 6 17 http://bmjopen.bmj.com/ 18 Trial design 8 Description of trial design including type of trial 7 19 20 (eg, parallel group, crossover, factorial, single 21 group), allocation ratio, and framework (eg, 22 superiority, equivalence, noninferiority, 23 exploratory) 24

25 on September 26, 2021 by guest. Protected copyright. 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 3 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Page 41 of 52 BMJ Open

1 2 3 4 5 Methods: Participants, interventions, 6 7 and outcomes 8 9 Study setting 9 Description of study settings (eg, community 7 10 clinic, academic hospital) and list of countries 11 where data will be collected. Reference to where 12 For peerlist of study review sites can be obtained only 13 14 Eligibility criteria 10 Inclusion and exclusion criteria for participants. If 7-8, Table 15 16 applicable, eligibility criteria for study centres

17 and individuals who will perform the http://bmjopen.bmj.com/ 18 interventions (eg, surgeons, psychotherapists) 19 20 Interventions 11a Interventions for each group with sufficient detail 8-9 21 to allow replication, including how and when they 22 23 will be administered 24

25 11b Criteria for discontinuing or modifying allocated on September 26, 2021 by guest. Protected copyright. 9 26 interventions for a given trial participant (eg, drug 27 dose change in response to harms, participant 28 request, or improving/worsening disease) 29 30 11c Strategies to improve adherence to intervention 8-9 31 32 protocols, and any procedures for monitoring 33 adherence (eg, drug tablet return, laboratory 34 tests) 35 36 11d Relevant concomitant care and interventions that 9 37 are permitted or prohibited during the trial 38 39 40 41 42 4 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open Page 42 of 52

1 2 3 4 Outcomes 12 Primary, secondary, and other outcomes, 9-10 5 including the specific measurement variable (eg, 6 systolic blood pressure), analysis metric (eg, 7 8 change from baseline, final value, time to event), 9 method of aggregation (eg, median, proportion), 10 and time point for each outcome. Explanation of 11 the clinical relevance of chosen efficacy and 12 For peerharm outcomes review is strongly recommended only 13 14 15 Participant timeline 13 Time schedule of enrolment, interventions 8-9 16 (including any run-ins and washouts), Figure 1

17 assessments, and visits for participants. A http://bmjopen.bmj.com/ 18 schematic diagram is highly recommended (see 19 20 Figure) 21 22 Sample size 14 Estimated number of participants needed to 10-11 23 achieve study objectives and how it was 24 determined, including clinical and statistical 25 assumptions supporting any sample size on September 26, 2021 by guest. Protected copyright. 26 27 calculations 28 29 Recruitment 15 Strategies for achieving adequate participant 7-8 30 enrolment to reach target sample size 31 32 33 34 35 36 37 38 39 40 41 42 5 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Page 43 of 52 BMJ Open

1 2 3 4 5 Methods: Assignment of 6 interventions (for controlled trials) 7 8 Allocation: 9 10 Sequence generation 16a Method of generating the allocation sequence 11-12 11 (eg, computer-generated random numbers), and 12 For peer review only 13 list of any factors for stratification. To reduce 14 predictability of a random sequence, details of 15 any planned restriction (eg, blocking) should be 16

17 provided in a separate document that is http://bmjopen.bmj.com/ 18 unavailable to those who enrol participants or 19 assign interventions 20 21 Allocation concealment mechanism 16b Mechanism of implementing the allocation 11-12 22 sequence (eg, central telephone; sequentially 23 24 numbered, opaque, sealed envelopes),

25 describing any steps to conceal the sequence on September 26, 2021 by guest. Protected copyright. 26 until interventions are assigned 27 28 Implementation 16c Who will generate the allocation sequence, who 11-12 29 will enrol participants, and who will assign 30 31 participants to interventions 32 33 Blinding (masking) 17a Who will be blinded after assignment to 11-12 34 interventions (eg, trial participants, care 35 providers, outcome assessors, data analysts), 36 and how 37 38 39 40 41 42 6 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open Page 44 of 52

1 2 3 4 17b If blinded, circumstances under which unblinding N/A 5 is permissible, and procedure for revealing a 6 participant’s allocated intervention during the trial 7 8 9 10 11 12 For peer review only 13 14 15 16

17 http://bmjopen.bmj.com/ 18 19 20 21 22 23 24

25 on September 26, 2021 by guest. Protected copyright. 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 7 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Page 45 of 52 BMJ Open

1 2 3 4 5 Methods: Data collection, 6 management, and analysis 7 8 Data collection methods 18a Plans for assessment and collection of outcome, 12 9 10 baseline, and other trial data, including any 11 related processes to promote data quality (eg, 12 For peerduplicate review measurements, training only of assessors) 13 and a description of study instruments (eg, 14 15 questionnaires, laboratory tests) along with their 16 reliability and validity, if known. Reference to

17 where data collection forms can be found, if nothttp://bmjopen.bmj.com/ 18 in the protocol 19 20 18b Plans to promote participant retention and 12 21 22 complete follow-up, including list of any outcome 23 data to be collected for participants who 24 discontinue or deviate from intervention 25 protocols on September 26, 2021 by guest. Protected copyright. 26 27 Data management 19 Plans for data entry, coding, security, and 12 28 29 storage, including any related processes to 30 promote data quality (eg, double data entry; 31 range checks for data values). Reference to 32 where details of data management procedures 33 34 can be found, if not in the protocol 35 36 Statistical methods 20a Statistical methods for analysing primary and 11 37 secondary outcomes. Reference to where other 38 details of the statistical analysis plan can be 39 found, if not in the protocol 40 41 42 8 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open Page 46 of 52

1 2 3 4 20b Methods for any additional analyses (eg, 11 5 subgroup and adjusted analyses) 6 7 20c Definition of analysis population relating to 11 8 protocol non-adherence (eg, as randomised 9 10 analysis), and any statistical methods to handle 11 missing data (eg, multiple imputation) 12 For peer review only 13 14 15 16

17 http://bmjopen.bmj.com/ 18 19 20 21 22 23 24

25 on September 26, 2021 by guest. Protected copyright. 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 9 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Page 47 of 52 BMJ Open

1 2 3 4 5 Methods: Monitoring 6 7 Data monitoring 21a Composition of data monitoring committee 12-13 8 (DMC); summary of its role and reporting 9 10 structure; statement of whether it is independent 11 from the sponsor and competing interests; and 12 For peerreference review to where further details only about its 13 charter can be found, if not in the protocol. 14 15 Alternatively, an explanation of why a DMC is 16 not needed

17 http://bmjopen.bmj.com/ 18 21b Description of any interim analyses and stopping 12-13 19 guidelines, including who will have access to 20 these interim results and make the final decision 21 22 to terminate the trial 23 24 Harms 22 Plans for collecting, assessing, reporting, and 12-13

25 managing solicited and spontaneously reported on September 26, 2021 by guest. Protected copyright. 26 adverse events and other unintended effects of 27 trial interventions or trial conduct 28 29 Auditing 23 Frequency and procedures for auditing trial 12-13 30 31 conduct, if any, and whether the process will be 32 independent from investigators and the sponsor 33 34 35 36 37 38 39 40 41 42 10 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open Page 48 of 52

1 2 3 4 5 Ethics and dissemination 6 7 Research ethics approval 24 Plans for seeking research ethics 14 8 9 committee/institutional review board (REC/IRB) 10 approval 11 12 Protocol amendments For25 peerPlans forreview communicating important only protocol 14 13 modifications (eg, changes to eligibility criteria, 14 outcomes, analyses) to relevant parties (eg, 15 16 investigators, REC/IRBs, trial participants, trial

17 registries, journals, regulators) http://bmjopen.bmj.com/ 18 19 Consent or assent 26a Who will obtain informed consent or assent from 14 20 potential trial participants or authorised 21 surrogates, and how (see Item 32) 22 23 26b Additional consent provisions for collection and N/A 24

25 use of participant data and biological specimens on September 26, 2021 by guest. Protected copyright. 26 in ancillary studies, if applicable 27 28 Confidentiality 27 How personal information about potential and 14 29 enrolled participants will be collected, shared, 30 and maintained in order to protect confidentiality 31 32 before, during, and after the trial 33 34 Declaration of interests 28 Financial and other competing interests for 18 35 principal investigators for the overall trial and 36 each study site 37 38 39 40 41 42 11 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Page 49 of 52 BMJ Open

1 2 3 4 Access to data 29 Statement of who will have access to the final 12 5 trial dataset, and disclosure of contractual 6 agreements that limit such access for 7 8 investigators 9 10 Ancillary and post-trial care 30 Provisions, if any, for ancillary and post-trial N/A 11 care, and for compensation to those who suffer 12 For peerharm from review trial participation only 13 14 Dissemination policy 31a Plans for investigators and sponsor to 14 15 communicate trial results to participants, 16

17 healthcare professionals, the public, and other http://bmjopen.bmj.com/ 18 relevant groups (eg, via publication, reporting in 19 results databases, or other data sharing 20 arrangements), including any publication 21 22 restrictions 23 24 31b Authorship eligibility guidelines and any intended N/A

25 use of professional writers on September 26, 2021 by guest. Protected copyright. 26 27 31c Plans, if any, for granting public access to the full 14 28 protocol, participant-level dataset, and statistical 29 code 30 31 32 33 34 35 36 37 38 39 40 41 42 12 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open Page 50 of 52

1 2 3 4 5 Appendices 6 7 Informed consent materials 32 Model consent form and other related Supplementary file 1 8 9 documentation given to participants and 10 authorised surrogates 11 12 Biological specimens For33 peerPlans forreview collection, laboratory onlyevaluation, and N/A 13 storage of biological specimens for genetic or 14 molecular analysis in the current trial and for 15 16 future use in ancillary studies, if applicable

17 http://bmjopen.bmj.com/ 18 19 *It is strongly recommended that this checklist be read in conjunction with the SPIRIT 2013 Explanation & Elaboration for important 20 clarification on the items. Amendments to the protocol should be tracked and dated. The SPIRIT checklist is copyrighted by the SPIRIT 21 Group under the Creative Commons “Attribution-NonCommercial-NoDerivs 3.0 Unported” license. 22 23 24

25 on September 26, 2021 by guest. Protected copyright. 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 13 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Page 51 of 52 BMJ Open

1 2 3 4 WHO Trial Registration Data Set 5 6 7 8 1. Primary Registry and Trial Identifying Number Clinicaltrial.gov 9 NCT04045093 10 11 2. Date of Registration in Primary Registry First posted on August 5, 2019 12 For peer review only 13 14 3. Secondary Identifying Numbers None 15 16 4. Source(s) of Monetary or Material Support None

17 http://bmjopen.bmj.com/ 18 5. Primary Sponsor None 19 20 6. Secondary Sponsor(s) None 21 22 7. Contact for Public Queries Prof Chung-Wah SIU, MD 23 24 Cardiology Division,

25 Department of Medicine, on September 26, 2021 by guest. Protected copyright. 26 The University of Hong Kong, 27 Hong Kong SAR, China 28 Tel: (852) 2255-4694, 29 Fax: (852) 2818-6304, 30 E-mail: [email protected] & [email protected]. 31 32 33 8. Contact for Scientific Queries Prof Chung-Wah SIU, MD 34 Cardiology Division, 35 Department of Medicine, 36 The University of Hong Kong, 37 Hong Kong SAR, China 38 Tel: (852) 2255-4694, 39 40 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open Page 52 of 52

1 2 3 Fax: (852) 2818-6304, 4 E-mail: [email protected] & [email protected]. 5 6 7 9. Public Title Protocol for Rationale and Design of DAbigatran for Stroke 8 PreVention In Atrial Fibrillation in MoDerate or Severe Mitral 9 Stenosis (DAVID-MS) study 10 11 10. Scientific Title Protocol for Rationale and Design of DAbigatran for Stroke 12 For peer reviewPreVention In Atrialonly Fibrillation in MoDerate or Severe Mitral 13 Stenosis (DAVID-MS) study 14 15 16 11. Countries of Recruitment Hong Kong (China) and China

17 http://bmjopen.bmj.com/ 18 12. Health Condition(s) or Problem(s) Studied Atrial fibrillation, mitral stenosis 19 20 13. Intervention Experimental Arm: 21 22 Dabigatran 150mg or Dabigatran 110mg (twice daily) according 23 to creatinine clearance level, twice daily) 24

25 on September 26, 2021 by guest. Protected copyright. 26 Active Comparator Arm: 27 28 Warfarin with dosage adjustment according to INR level 29 (targeting to INR 2-3) 30 31 14. Key Inclusion and Exclusion Criteria Inclusion criteria: 32 33  Patients with atrial fibrillation documented with standard 34 12-lead ECG documented atrial fibrillation on the day of 35 screening or randomization 36  Patients with age 18 years old or above 37 38  Patients with moderate or severe mitral stenosis, i.e. 39 mitral valvular area (MVA) <1.5cm2 40 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Page 53 of 52 BMJ Open

1 2 3  Patients should be able to provide a written informed 4 5 consent 6  Patients should have all 4 inclusion-criteria fulfilled to be 7 qualified for the study 8 9 Exclusion criteria: 10  Patients with prosthetic valve, or with active endocarditis 11  Patients with planned valvular intervention within 1 year 12 For peer review only 13  Patients with left atrial appendage occlusive device 14  Patients with planned AF ablation 15  Patients with history of intracranial, intraocular, spinal, or 16 retroperitoneal bleeding 17 http://bmjopen.bmj.com/ 18  Unexplained anemia (haemoglobin level <10g/dL) or 19 thrombocytopenia (platelet count <100x10*9/L) 20  Need for anticoagulant therapy of disorders other than 21 atrial fibrillation 22  Patients receiving antiplatelet therapy for disorders other 23 an atrial fibrillation 24

25  Uncontrolled hypertension on September 26, 2021 by guest. Protected copyright. (systolic blood pressure 26 >180mmHg and/or diastolic blood pressure >100mmHg) 27  Estimated creatinine clearance equal to or less than 28 30mL/min 29  Liver dysfunction of Child Pugh stage B or C 30 31  Women who are pregnant or of childbearing potential 32 who refuse to use a medically acceptable form of 33 contraception throughout the study 34  Patients considered unreliable by the investigator or 35 have a life expectancy less than 1 year because of 36 concomitant disease, or has any condition, which in the 37 38 opinion of the investigator, would not allow safe 39 40 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open Page 54 of 52

1 2 3 participation in the study (e.g. drug addiction, alcohol 4 abuse) 5 6 7 15. Study Type Study Type: Interventional (Clinical Trial) 8 Participants Allocation: Randomized (details in protocol 9 manuscript) 10 11 Intervention Model: Parallel Assignment 12 For peer review only 13 Masking: None (Open Label) 14 Primary Purpose: Prevention 15 16

17 16. Date of First Enrollment June 1, 2020 http://bmjopen.bmj.com/ 18 19 17. Target Sample Size 686 20 21 18. Recruitment Status Pending 22 23 19. Primary Outcome(s) 1. Stroke, time frame: 1 year 24

25 2. Systemic embolism, time frame: on September 26, 2021 by guest. Protected copyright. 1 year 26 27 20. Key Secondary Outcome(s) 1. Ischemic stroke, time frame: 1 year 28 29 2. Hemorrhagic stroke, time frame: 1 year 30 3. Intracranial haemorrhage, time frame: 1 year 31 32 4. Major bleeding, time frame: 1 year 33 34 5. Death, time frame: 1 year 35 36 37 38 39 40 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Protocol for Rationale and Design of DAbigatran for Stroke PreVention In Atrial Fibrillation in MoDerate or Severe Mitral Stenosis (DAVID-MS) : A Randomized, Open-label study For peer review only Journal: BMJ Open

Manuscript ID bmjopen-2020-038194.R2

Article Type: Protocol

Date Submitted by the 21-Jun-2020 Author:

Complete List of Authors: zhou, Mi; University of Hong Kong, Department of Cardiology Chan, Esther; University of Hong Kong, Department of Cardiology Hai, Jo Jo; University of Hong Kong, Department of Cardiology Wong, Chun Ka; University of Hong Kong, Department of Cardiology LAU, Yuk-Ming; University of Hong Kong Huang, Duo; University of Hong Kong, Department of Cardiology LAM, Cheung-Chi; University of Hong Kong TAM, Chor-Cheung; University of Hong Kong WONG, Anthony; Queen Mary Hospital, University of Hong Kong, Medicine

LAM, Ho ; Tuen Mun Hospital, Medicine and Geriatrics on September 26, 2021 by guest. Protected copyright. NG, Andrew; Grantham Hospital, Cardiac Medical Unit FAN, Katherine; Grantham Hospital, Cardiac Medical Unit JIM, Man-Hong; Grantham Hospital, Cardiac Medical Unit Kai Hang, Yiu; University of Hong Kong, Department of Cardiology Yan, BP ; Chinese University of Hong Kong, Medicine & Therapeutics SIU, Chung-Wah; University of Hong Kong, Department of Cardiology

Primary Subject Cardiovascular medicine Heading:

Secondary Subject Heading: Pharmacology and therapeutics

Valvular heart disease < CARDIOLOGY, CLINICAL PHARMACOLOGY, Keywords: Adult cardiology < CARDIOLOGY

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 Protocol for Rationale and Design of 4 5 6 DAbigatran for Stroke PreVention In Atrial Fibrillation in 7 8 MoDerate or Severe Mitral Stenosis (DAVID-MS): A Randomized, 9 10 Open-label study 11 12 13 1Mi ZHOU, MBBS; 2Esther W. CHAN, PhD; 1Jojo HAI MBBS; 1Chun-Ka WONG, 14 MBBS; 1Yuk-Ming LAU, MBBS; 1Duo HUANG, MBBS, PhD; 1Cheung-Chi LAM, 15 MBBS; 1Chor-Cheung TAM, MBBS; 1 Anthony YT WONG, MBBS; 1 Arthur SY 16 For1 peer review3 only 3 17 YUNG, MBBS; Kelvin KW CHAN, MBBS; Yingqing FENG, MD; Ning TAN, 3 4 4 5 18 MD; Ji-Yan CHEN, MD; Chi-Yui YUNG, MBBS; Kwok-Lun LEE, MBBS; Chun- 19 Wai CHOI, MB ChB; 5Ho LAM, MBBS; 6Andrew NG, MBBS; 6Katherine FAN, 20 MBBS; 6Man-Hong JIM, MD; 1Kai-Hang YIU, MD, PhD; 7Bryan P. YAN MD;# and 21 1Chung-Wah SIU, MD.# 22 23 1Cardiology Division, Department of Medicine, Queen Mary Hospital, the 24 2 25 University of Hong Kong, Hong Kong SAR, China; Centre for Safe Medication 26 Practice and Research, Department of Pharmacology and Pharmacy, Li Ka Shing 27 Faculty of Medicine, the University of Hong Kong, Hong Kong SAR, China; 28 3Department of Cardiology, Guangdong Cardiovascular Institute, Guangdong 29 Provincial Key Laboratory of Coronary Heart Disease Prevention, Guangdong 30 General Hospital, Guangdong Academic of Medical Sciences, Guangzhou, China; 31 4Department of Cardiology, Ruttonjee and Tang Siu Kin Hospital, Hong Kong SAR, 32 China; 5Department of Medicine and Geriatrics, Tuen Mun Hospital, Hong Kong

33 http://bmjopen.bmj.com/ 6 34 SAR, China; Cardiac Medical Unit, The Grantham Hospital, Hong Kong SAR, 35 China; and 7Division of Cardiology, Department of Medicine and Therapeutics, The 36 Chinese University of Hong Kong, Hong Kong SAR, China. 37 38 #These authors are co-corresponding authors. 39 40

41 Protocol Version: 1 on September 26, 2021 by guest. Protected copyright. 42 43 Protocol Date: March 10, 2020 44 Wordcounts: 3452 45 Keywords: Mitral stenosis, atrial fibrillation, non-vitamin K oral anticoagulant, 46 47 warfarin 48 49 Correspondence: 50 Chung-Wah SIU, MD 51 52 Cardiology Division, 53 Department of Medicine, 54 The University of Hong Kong, 55 56 Hong Kong SAR, China 57 58 59 1 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 Tel: (852) 2255-4694, 4 5 Fax: (852) 2818-6304, 6 E-mail: [email protected] & [email protected]. 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 2 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 ABSTRACT 4 5 6 Introduction: Current international guidelines recommend non-vitamin K oral 7 8 9 anticoagulants (NOACs) for stroke prevention amongst patients with non-valvular 10 11 atrial fibrillation (AF) at significant ischaemic stroke risk given the superior safety 12 13 and comparable efficacy of NOACs over warfarin. Nonetheless, the safety and 14 15 16 effectiveness For of NOACs peer have not review been evaluated only in patients with AF with 17 18 underlying moderate or severe mitral stenosis (MS), hence the recommended 19 20 stroke prevention strategy remains warfarin therapy. 21 22 23 24 25 Method and analysis: MS remain disproportionately prevalent in Asian countries 26 27 compared with the developed countries. This prospective, randomized, open-label 28 29 trial with blinded endpoint adjudication aims to evaluate the safety and efficacy of 30 31 32 dabigatran for stroke prevention in AF patients with moderate or severe MS.

33 http://bmjopen.bmj.com/ 34 Patients with AF aged ≥18 years with moderate or severe mitral stenosis not 35 36 planned for valvular intervention in the coming 12 months will be randomized in a 37 38 39 1:1 ratio to receive dabigatran 110 mg or 150 mg twice daily or warfarin with INR 40

41 2-3 in an open-label design. Patients with estimated creatinine clearance <30 on September 26, 2021 by guest. Protected copyright. 42 43 ml/min, or with a concomitant indication for anti-platelet therapy will be excluded. 44 45 46 The primary outcome is a composite of stroke and systemic embolism. Secondary 47 48 outcomes are ischaemic stroke, systemic embolism, hemorrhagic stroke, 49 50 intracranial haemorrhage, major bleeding, and death. The estimated required 51 52 sample size is approximately 686 participants. 53 54 55 56 57 58 59 3 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 Ethics and dissemination: The study protocol has been approved by the 4 5 6 Institutional Review Board of The University of Hong Kong, and Hong Kong West 7 8 Cluster, Hospital Authority, Hong Kong for Fung Yiu King Hospital, Grantham 9 10 11 Hospital, Queen Mary Hospital and Tung Wah Hospital in Hong Kong. Results will 12 13 14 be published in peer-reviewed journals. 15 16 For peer review only 17 18 19 Registration details: ClinicalTrials.gov (NCT04045093). 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 4 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 STRENGTHS AND LIMITATIONS OF THIS STUDY 4 5 6  This is the first investigator-initiated, open-label, randomized clinical trial to 7 8 compare effectiveness and safety of dabigatran and warfarin therapy for 9 10 stroke prevention in patients with atrial fibrillation (AF) and moderate or 11 12 13 severe mitral stenosis (MS). 14 15  The study is designed to provide clinicians with robust, much-needed 16 For peer review only 17 information regarding stroke prevention strategy for patients with AF and 18 19 20 moderate or severe MS. 21 22  The results of study will have immediate and long-term impacts on the 23 24 management of these very high-risk patients with AF. 25 26 27  This study will provide the necessary evidence for establishing international 28 29 clinical practice guidelines for stroke prevention in patients with AF and MS. 30 31  Since the clinical trial will be conducted mainly in Hong Kong and mainland 32

33 http://bmjopen.bmj.com/ 34 China, it is expected that most recruited subjects will be of Chinese ethnicity, 35 36 which may limit generalizability of the trial results. 37 38 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 5 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 INTRODUCTION 4 5 6 Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia 7 8 encountered in clinical practice.1 Patients with AF are at increased risk of 9 10 ischaemic stroke and systemic thromboembolism due to the formation and 11 12 embolism of left atrial thrombi, 1-3 hence long-term oral anticoagulant (OAC) for 13 14 15 thromboprophylaxis is the cornerstone in AF management. In previous 16 For peer review only 17 randomized clinical trials in the last century, warfarin has been shown to be highly 18 19 effective in reducing stroke risk compared with placebo by as much as 64% in 20 21 4 22 patients with AF. More recently, non-vitamin K oral anticoagulants (NOACs) have 23 24 consistently demonstrated to be safer and more effective for stroke prevention in 25 26 patients with non-valvular AF compared with warfarin and have become the 27 28 29 recommended standard of care for the management of stroke prevention in non- 30 31 valvular AF. 32

33 http://bmjopen.bmj.com/ 34 35 While the stroke risk amongst patients with AF appears heterogeneous,5 6 patients 36 37 38 with underlying valvular heart disease, particularly mitral stenosis (MS) are at very 39 40 high risk for stroke with an annual stroke risk ranging from 4% to 17% if left un-

41 on September 26, 2021 by guest. Protected copyright. 42 anticoagulated7 and the highest recurrences.8 However, patients with AF and 43 44 9 45 underlying MS are typically excluded in randomized control trials. As a result, 46 47 current international guidelines for management of AF do not recommend NOACs 48 49 for stroke prevention in patients with AF and underlying moderate or severe MS.3 50 51 52 Nonetheless, off-label use of NOAC in patients with AF and MS is not uncommon 53 54 in the real-world practice. In a recently published retrospective, observational 55 56 57 58 59 6 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 analysis from the Republic of Korea,10 in a cohort of 7,357 patients with MS 4 5 6 receiving anticoagulation therapy, 35% of these patients were in fact treated with 7 8 NOAC with the remaining 65% with warfarin. More importantly, after propensity 9 10 matching, it was shown that patients treated with NOAC had a substantially lower 11 12 risk of ischaemic stroke/systemic embolism with an annualized risk of 2.22%/year, 13 14 15 compared to that of 4.19%/year for patients treated with warfarin, (adjusted HR: 16 For peer review only 17 0.28; 95% confidence interval (CI): 0.18 to 0.45), suggesting a potential role of 18 19 NOAC amongst patients with AF and underlying MS.8 10 11 20 21 22 23 24 This is of particular importance for Asian AF patients, in whom MS remains 25 26 relatively prevalent despite a declining trend.7 More importantly, the much higher 27 28 29 baseline risk of intracranial haemorrhage and apparently higher ischaemic stroke 30 31 risk in Asian populations potentially undermines the benefits of warfarin 32 12,13 14 33 therapy. Notably, compared with warfarin, the effectiveness, and safety of http://bmjopen.bmj.com/ 34 35 NOACs appear to be even more superior in Asian populations than Caucasian 36 37 38 populations as shown in sub-analyses of pivotal randomized controls trials15-17 as 39 40 well as in studies using real-world data.18-23 To our knowledge, this is the first

41 on September 26, 2021 by guest. Protected copyright. 42 multicentre randomized control trial aims to comparing NOAC to warfarin to 43 44 45 address the knowledge in stroke prevention strategy in patients with AF and 46 47 moderate or severe MS. This will have immediate and long-term impacts on the 48 49 management of these very high-risk patients with AF. 50 51 52 53 54 55 56 57 58 59 7 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 METHODS AND ANALYSIS 4 5 6 This clinical trial protocol follows the Standard Protocol Items: Recommendations 7 8 9 24 25 10 for Interventional Trials (SPIRIT). The underlying protocol follows the 11 12 13 Consolidated Standards of Reporting Trials (CONSORT). 26 27 The study is 14 15 16 registered withFor the www.ClinicalTrials.gov peer review (NCT04045093). only 17 18 19 20 21 Study Design 22 23 This is an investigator-initiated, open-label, randomized clinical trial to compare 24 25 effectiveness and safety of dabigatran 150 mg or 110 mg twice daily according to 26 27 28 kidney function with warfarin therapy with target international normalized ratio 29 30 (INR) 2-3 for stroke prevention in patients with AF and moderate or severe MS. 31 32

33 http://bmjopen.bmj.com/ 34 Study Participants 35 36 37 Patients will be recruited from participating specialist cardiology centres in Hong 38 39 Kong SAR China and Mainland China. Written informed consent will be obtained 40

41 from all study participants. Table 1 summarizes the inclusion and exclusion criteria on September 26, 2021 by guest. Protected copyright. 42 43 44 for the study. In brief, patients with no symptoms, aged 18 years or above will be 45 46 eligible if they have AF documented on standard 12-lead electrocardiography 47 48 (ECG) performed at screening or randomization and moderate or severe MS as 49 50 2 2 51 defined as the mitral valvular area (MVA) of 1.0-1.5 cm and <1.0 cm , respectively. 52 53 Reasons for exclusion include the presence of symptoms, prosthetic valve, left 54 55 atrial appendage occlusive device, and/or active endocarditis; planned valvular 56 57 58 59 8 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 intervention and/or planned AF ablation; history of major bleeding including 4 5 6 intracranial, intraocular, spinal or retroperitoneal haemorrhage; unexplained 7 8 anaemia with haemoglobin level <10 g/dL, or thrombocytopenia with platelet count 9 10 <100×109/L; need for anticoagulant or antiplatelet therapy of conditions other than 11 12 AF; concomitant use of potent P-gp inhibitor(s) or drugs with a known interaction 13 14 15 with dabigatran; uncontrolled hypertension; significant kidney impairment with 16 For peer review only 17 estimated creatinine clearance (CrCl) ≤30 mL/min by the Cockcroft-Gault Formula; 18 19 28 liver dysfunction of Child-Pugh Stage B or C; 29 pregnancy or if there is child- 20 21 22 baring potential during the full duration of the study. In addition, patients 23 24 considered unsuitable by the investigator including short life expectancy <1 year 25 26 due to concomitant disease, substance and/or alcohol abuse or other medical 27 28 29 conditions. 30 31 32

33 Study Procedures http://bmjopen.bmj.com/ 34 35 After providing written informed consent, all study participants will be randomly 36 37 38 assigned to receive dabigatran or to receive warfarin. The procedure of the trial is 39 40 summarized in figure 1, the trial will primarily be conducted in Hong Kong and

41 on September 26, 2021 by guest. Protected copyright. 42 Mainland China. In Hong Kong, there is no local guideline on dabigatran dosage 43 44 45 in relation to renal function. In Mainland China, dosage reduction to 110mg two 46 47 times per day was recommended in patients with creatinine clearance in the range 48 49 30-49 ml/min. For patients randomized to receive dabigatran, the dosage regimen 50 51 52 will be determined according to the respective estimated CrCl or if concomitantly 53 54 taking interacting drugs requiring dosage adjustment. Patients with estimated CrCl 55 56 57 58 59 9 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 above 50 ml/min will receive dabigatran 150 mg twice daily, whereas those with 4 5 6 CrCl between 30 to 50 ml/min will receive dabigatran 110 mg twice daily. For 7 8 patients previously on warfarin randomized to receive dabigatran, dabigatran will 9 10 initiate after discontinuation of warfarin with an INR less than or equal to 2. At the 11 12 end of study, patients randomized to dabigatran will be switched back to warfarin. 13 14 15 Warfarin will be initiated 3 days prior to the termination of dabigatran. INR will be 16 For peer review only 17 checked 5 days after initiation of warfarin i.e., 2 days after termination of dabigatran 18 19 to minimise the potential impact of remaining dabigatran levels in elevating the 20 21 22 INR. On the other hand, for those randomized to receive warfarin, INR will be 23 24 measured at least every 8 weeks with a target INR of 2.0 to 3.0. The time in 25 26 therapeutic range (TTR) will be calculated for each study participant using 27 28 30 29 Rosendaal method, in which INR will be assumed to change in a linear manner 30 31 between measurements, and INR values on the days without measurement are 32

33 interpolated. The percentage of time during which a study participant has an INR http://bmjopen.bmj.com/ 34 35 within 2.0-3.0 is taken as TTR. The first follow-up visit will be scheduled 14 days 36 37 38 after randomization and then every 4 months during the study period of 1 year 39 40 (Table 2). Criteria for discontinuation or change of allocated treatment include

41 on September 26, 2021 by guest. Protected copyright. 42 patient request, drug allergy, intolerable adverse drug reaction and development 43 44 45 of other contraindication. Patients who are randomized to receive dabigatran would 46 47 be switched to warfarin if CrCl is below 30 ml/min and/or develop liver dysfunction 48 49 of Child-Pugh Stage B or C. 50 51 52 53 54 55 56 57 58 59 10 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 4 5 6 Outcomes 7 8 The primary outcome is a composite of stroke or systemic embolism at 1 year. 9 10 Secondary outcomes are ischaemic stroke, systemic embolism, hemorrhagic 11 12 stroke, intracranial haemorrhage, major bleeding and death at 1 year. Stroke is 13 14 15 defined as a neurological deficit of sudden onset that persisted for more than 24 16 For peer review only 17 hours and corresponded to a vascular territory that cannot be explained by other 18 19 causes (such as trauma, infection or vasculitis). Stroke will be further classified as 20 21 22 ischaemic stroke and hemorrhagic stroke according to computerized axial 23 24 tomography or magnetic resonance imaging of the brain. Intracranial haemorrhage 25 26 (ICH) consists of hemorrhagic stroke (intra-cerebral haemorrhage and cerebellar 27 28 29 haemorrhage ), subdural haemorrhage , and subarachnoid haemorrhage , and will 30 31 be confirmed with computerized axial tomography or magnetic resonance imaging 32

33 of the brain. Systemic embolism is defined as an acute vascular occlusion of an http://bmjopen.bmj.com/ 34 35 extremity or organ other than the brain, documented by imaging, surgery, and/or 36 37 38 autopsy. 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 Major bleeding is defined as a drop in the haemoglobin level of at least 2 g/dL, 43 44 45 transfusion of at least 2 units of blood, or symptomatic bleeding in a critical area 46 47 or organ. Life-threatening bleeding includes fatal bleeding, symptomatic 48 49 intracranial bleeding, bleeding with a haemoglobin drop of at least 5 g/dL, or 50 51 52 bleeding requiring transfusion of at least 4 units of blood or inotropic agents or 53 54 55 56 57 58 59 11 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 requiring surgery. All outcomes will be adjudicated by 2 independent investigators 4 5 6 in a blinded fashion. 7 8 9 10 Sample Size Calculation 11 12 The primary analysis is to test whether dabigatran is noninferior to warfarin for 13 14 15 ischaemic stroke prevention in patients with AF and moderate or severe MS. The 16 For peer review only 17 potential for dabigatran to preserve at least 50% of the effectiveness of warfarin is 18 19 considered clinically meaningful, as noninferior in patients with AF and moderate 20 21 22 or severe MS. The noninferiority margin is 1.49, which is derived from the only 23 24 observation study comparing vitamin K antagonist with NOAC in patients with AF 25 26 and MS.10 In the study, the annual ischaemic stroke risk of patients with AF and 27 28 29 MS receiving vitamin K antagonist and NOAC are 4.19%/year and 2.22%/year 30 31 respectively. Accordingly, based on the margin of error (4.66%) and the current 32

33 population of Hong Kong (7,500,700), a sample size of 686 patients (343 patients http://bmjopen.bmj.com/ 34 35 in the vitamin K antagonist group and 343 in the dabigatran group) including 10% 36 37 38 attrition would be needed to satisfy the noninferiority hypothesis with the upper 39 40 boundary of the one-sided 95% confidence interval (CI) (or equivalent with a 90%

41 on September 26, 2021 by guest. Protected copyright. 42 two-sided CI) and the Hazard ratio (HR) of the primary outcome below the 43 44 45 noninferiority margin of 1.49. Hierarchical analysis for superiority will be performed 46 47 if noninferiority is established. 48 49 50 51 52 Statistical Analysis 53 54 55 56 57 58 59 12 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 Baseline data will be reported as means and standard deviations for continuous 4 5 6 data and as numbers and percentages for categorical data. All endpoints will be 7 8 analysed according to the intention-to-treat principle, with all patients who undergo 9 10 randomization included in the analysis. Clinical events that occur after 11 12 randomization and until the end of the study (at 1 year or mortality) will be included 13 14 15 in the primary analysis of clinical outcomes. A p-value <0.05 considered as 16 For peer review only 17 significant. Calculations will be performed using SPSS software (version 12.0). 18 19 20 21 22 Randomization 23 24 Randomization will be stratified to each study site to account for variations in 25 26 patient demographics and diagnoses. At each site, patients will be randomised to 27 28 29 “permuted blocks of four” (two of each study arm) to assist in equality of numbers 30 31 in each arm. An independent research officer who are blinded to this study will 32

33 http://bmjopen.bmj.com/ 34 generate the random-number table. Study staff responsible for enrolment will be 35 36 37 informed of randomization assignment by phone. Subjects and clinicians will not 38 39 40 be blinded to the randomization assignment. Data staff responsible for data entry

41 on September 26, 2021 by guest. Protected copyright. 42 will be blinded from randomization assignment. 43 44 45 46 47 Data collection and management 48 49 50 After enrolment, each subject will be assigned a unique identifier to be used in 51 52 53 database. Data will be entered by study staff and data accuracy will be verified by 54 55 56 57 58 59 13 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 4 study principal investigator. Data quality control measures include queries to 5 6 7 identify missing data, outliers and discrepancies. The database will be password 8 9 10 11 protected and encrypted. Only study staff will have access to the database. All 12 13 14 paper records will be deidentified and stored securely in a locked cabinet for 5 15 16 years. SubjectsFor who withdrawpeer from review the study will only have continuous monitoring 17 18 19 20 stopped, usual care continued and final outcome collected for analysis. 21 22 23 24 25 Data monitoring and safety 26 27 28 An independent Safety Committee will be established comprising of an Emergency 29 30 Clinician, Clinical Pharmacologist and Toxicologist. They will receive regular 31 32 reports during patient enrolment and be notified of any adverse drug reaction and

33 http://bmjopen.bmj.com/ 34 35 study protocol violation. The Safety Committee is led by Professor Bernard 36 37 Cheung from the Clinical Pharmacology, Faculty of Medicine, the University of 38 39 Hong Kong, Hong Kong SAR, China. For patient safety, discontinuation of the 40

41 study is at the discretion of the cardiology clinician to enable informed decisions to on September 26, 2021 by guest. Protected copyright. 42 43 44 be made regarding subsequent management and alternative medication use. Any 45 46 medication or therapy, intervention or procedure thought to be necessary for the 47 48 safe management of the patient may be administered at the discretion of the 49 50 51 managing clinician. 52 53 54 55 Patient and public involvement 56 57 58 59 14 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 4 We received input from clinicians and patients which guided the design of the 5 6 7 current study and choice of research questions. No patients were directly involved 8 9 10 11 in the design of the study and choice of outcome measures. No patients will be 12 13 14 involved in recruitment or conduct of the study. Results of the study will be 15 16 For peer review only 17 18 disseminated to subjects, the public and the scientific community. 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 15 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 4 ETHICS AND DISSEMINATION 5 6 7 This research protocol complies with the Declaration of Helsinki and the 8 9 International Conference on Harmonisation-Good Clinical Practice. The study 10 11 protocol has been approved by the Institutional Review Board of The University of 12 13 14 Hong Kong, and Hong Kong West Cluster, Hospital Authority, Hong Kong for Fung 15 16 Yiu King Hospital,For Grantham peer Hospital, review Queen Mary only Hospital and Tung Wah 17 18 Hospital in Hong Kong. Written informed consents will be obtained from all study 19 20 21 participants by study staff responsible for recruitment (Supplementary File 1). 22 23 Important protocol modifications will be conveyed to investigators, Institutional 24 25 26 27 Review Board, trial registries, regulators, journals and trial participants. After 28 29 30 enrolment, each subject will be assigned a unique identifier to be used in database. 31 32

33 http://bmjopen.bmj.com/ 34 Personal identity of subjects will not be used for any public purpose, publication, 35 36 37 or transmitted outside of the study team. 38 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 43 Dataset used during the study will be available from the corresponding author on 44 45 46 47 reasonable request. Collaboration with other investigators will be welcomed. The 48 49 50 results of the trial will be published in peer-reviewed journals and presented in 51 52 53 conferences. 54 55 56 57 58 59 16 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 17 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 DISCUSSION 4 5 6 To our knowledge, this is the first study to evaluate dabigatran as an alternative to 7 8 warfarin for stroke prevention in patients with AF and moderate or severe MS. The 9 10 results could fill the gap in stroke prevention strategy for this specific group of 11 12 patients with AF with immediate and long-term impacts on clinical practice. 13 14 15 16 For peer review only 17 AF is the most commonly encountered sustained cardiac arrhythmia in clinical 18 19 practice.1 31 While patients with AF have in general increased risk of stroke, 4 20 21 22 subgroups are at particularly high risk necessitating long-term anticoagulation 23 24 therapy. These include (1) patients with non-valvular AF and high CHA2DS2-VASc 25 26 score, (2) patients with AF and hypertrophic cardiomyopathy, (3) patients with AF 27 28 29 and MS, and (4) patients with AF and mechanical heart valves. (Figure 2) In the 30 31 past decade, NOAC has emerged as the preferred agent over warfarin for stroke 32

33 http://bmjopen.bmj.com/ prevention in patients with non-valvular AF and high CHA2DS2-VASc score. In the 34 35 4 pivotal studies comparing NOACs and warfarin in patients with non-valvular AF, 36 37 38 NOACs are at least as effective as warfarin to reduce stroke and at the same time 39 40 are much safer alternatives in terms of life-threatening bleeding complications.

41 on September 26, 2021 by guest. Protected copyright. 42 Nonetheless, direct extrapolation of these results to other subgroups of AF patients 43 44 45 may not be appropriate, given the different mechanisms of thrombus formation in 46 47 different diseases. For instance, in the RE-ALIGN trial randomizing 252 patients 48 49 with recent mechanical valvular replacement in a 2:1 ratio to receive either 50 51 52 dabigatran or warfarin, there was an excess of thromboembolic as well as bleeding 53 54 55 56 57 58 59 18 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 events among patients randomized to dabigatran, rendering the study prematurely 4 5 32 6 terminated. 7 8 9 10 On the other hand, international guidelines do not recommend NOACs for patients 11 12 with AF and moderate or severe mitral stenosis due to the lack of reliable data 13 14 15 from clinical trials. Nonetheless, it remains undetermined whether NOACs can be 16 For peer review only 17 used as an alternative to warfarin for patients with AF and moderate or severe MS 18 19 due to the lack of clinical trial. The current study has several important implications, 20 21 22 particularly in Asian countries. First, while MS is now a rare condition in developed 23 24 countries, it remains relatively prevalent in many Asian countries. In addition, the 25 26 risk of stroke amongst patients with AF and MS is only second to those with 27 28 29 mechanical valvular replacement ranging from 4 to 17%. Second, previous 30 31 epidemiological studies12,13 14 21 33 34 and sub-analyses of the pivotal NOAC trials15- 32 17 33 have consistently reported a much higher nominal risk of ICH amongst Asians http://bmjopen.bmj.com/ 34 35 than non-Asians, favouring NOACs over warfarin therapy. More importantly, the 36 37 38 notoriously poor time in therapeutic range (TTR) for warfarin in Asian populations 39 40 observed in real-world data18 34-38 and pivotal NOAC trials15-17 substantially

41 on September 26, 2021 by guest. Protected copyright. 42 undermines the overall clinical benefits of warfarin therapy. In fact, the annual 43 44 45 incidence of ICH amongst patients with AF and MS treated with warfarin has been 46 47 reported to be as high as 0.93% per year,10 urging a much safer alternative. 48 49 50 51 52 In the present study, the NOAC of choice is dabigatran, the first NOAC with an 53 54 approved indication for stroke prevention in non-valvular AF patients from the 55 56 57 58 59 19 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 United States Food and Drug Administration in 2009. In the pivotal study, the RE- 4 5 39 6 LY study, patients with non-valvular AF with CHADS2>1 were randomly assigned 7 8 to two doses of dabigatran: 110 mg or 150 mg twice daily; or adjusted-dose 9 10 warfarin. After a median follow-up of 2.0 years, the low-dose regime was found to 11 12 be as effective as warfarin in preventing the primary endpoint (a composite of 13 14 15 stroke and systemic embolism) (1.52%/year vs. 1.69%/year), but with a 16 For peer review only 17 substantially lower risk of major bleeding and ICH.39 On the other hand, the 18 19 standard dose dabigatran (150mg BD) is superior to warfarin in reducing the 20 21 39 22 primary composite endpoint and ICH, with a comparable risk of major bleeding. 23 24 In an analysis comparing the effectiveness and safety of dabigatran according to 25 26 the ethnicity of study participants, dabigatran appears to be more effective in stroke 27 28 29 prevention as well as safer in terms of ICH compared with warfarin. This is in 30 31 concordance to subsequent real-world cohorts of AF patients from territory-wide 32

33 registries from Asia Pacific region. Plausible explanations include suboptimal http://bmjopen.bmj.com/ 34 35 quality of warfarin therapy with low time in therapeutic range and a higher risk of 36 37 38 ICH in Asian populations.18 23 35 40 41 An additional reason for the choice of 39 40 dabigatran in the present study is the wide availability of its antidote, idarucizumab

41 on September 26, 2021 by guest. Protected copyright. 42 in Asian countries, which provides extra-protection of patients in the clinical trial. 43 44 45 46 47 The study is designed to provide clinicians with robust, much-needed information 48 49 regarding stroke prevention strategy for patients with AF and moderate or severe 50 51 52 MS. The results will have immediate and long-term impacts on the management 53 54 of these very high-risk patients with AF. 55 56 57 58 59 20 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 Author contributions: 4 5 6 MZ, EWC, JH, CKW, BPY and CWS contributed to the conception and design of 7 8 the study. MZ, EWC, JH, CKW, YML, DH, CCL, CCT, AYTW, ASYY, KKWC, YF, 9 10 NT, JYC, CYY, KLL, CWC, HL, AN, KF, MHJ, KHY, BPY and CWS contributed to 11 12 13 the acquisition of data. Data analysis and interpretation will be conducted by MZ, 14 15 16 EWC, JH, CKW,For BPY andpeer CWS. MZ, review EWC, JH, CKW, only BPY and CWS wrote first 17 18 19 20 draft of the protocol and revised the protocol critically for important intellectual 21 22 23 content. All authors have read and approved the final version of the manuscript to 24 25 26 27 be published. 28 29 30 31 32 Acknowledgement:

33 http://bmjopen.bmj.com/ 34 35 None. 36 37 38 39 Funding statement: 40

41 None. on September 26, 2021 by guest. Protected copyright. 42 43 44 45 46 Conflict of interest: 47 48 None declared. 49 50 51 52 53 54 55 56 57 58 59 21 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 Legends: 4 5 6 Figure 1. Design of the DAVID-MS study. 7 8 Figure 2. Four main groups of patients with AF requiring long-term 9 10 anticoagulation therapy. MS: mitral stenosis; NOAC: non-vitamin K oral 11 12 anticoagulant; and VKA: vitamin K antagonist. 13 14 15 16 SupplementaryFor Files: peer review only 17 18 19 20 Supplementary File 1: Model consent form in English 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 22 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 Tables: 4 5 6 Table 1. Inclusion and Exclusion Criteria 7 8 Inclusion Criteria 9 10  Patients with AF documented with standard 12-lead ECG documented AF on 11 12 the day of screening or randomization 13  Patients with age >18 years 14 15  Patients with moderate or severe MS i.e., MVA <1.5 cm2 16  PatientsFor should bepeer able to provide review a written, informed only consent. 17 18  Patients should have all 4 inclusion-criteria fulfilled to be qualified for the 19 20 study. 21 Exclusion criteria 22 23  Patients with prosthetic valve, or with active endocarditis 24 25  Patients with heart failure symptom 26  Patients with planned valvular intervention within 1 year 27 28  Patients with left atrial appendage occlusive device 29 30  Patients with planned AF ablation 31  Patients with a history of intracranial, intraocular, spinal, or retroperitoneal 32

33 bleeding http://bmjopen.bmj.com/ 34 35  Unexplained anemia (haemoglobin level <10 g/dL) or thrombocytopenia 36 (platelet count <100×109/L) 37 38  Need for anticoagulant therapy of disorders other than AF 39  Patients receiving antiplatelet therapy for disorders other than AF 40

41  Patients receiving concomitant P-gp inhibitors and/or medications known to on September 26, 2021 by guest. Protected copyright. 42 43 interact with dabigatran 44  Uncontrolled hypertension (systolic blood pressure >180 mmHg and/or 45 46 diastolic blood pressure >100 mmHg) 47 48  Estimated creatinine clearance ≤30 mL/min 49  Liver dysfunction of Child-Pugh stage B or C 50 51  Women who are pregnant or of childbearing potential who refuse to use a 52 medically acceptable form of contraception throughout the study 53 54 55 56 57 58 59 23 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3  Patients considered unreliable by the investigator or have a life expectancy 4 5 less than 1 year because of concomitant disease, or has any condition, which 6 in the opinion of the investigator, would not allow safe participation in the 7 8 study (e.g. drug addiction, alcohol abuse). 9 10 Abbreviations: AF: atrial fibrillation; MS: mitral stenosis; MVA: mitral valvular area 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 24 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Page 27 of 55 BMJ Open

1 2 3 4 5 Table 2. Study visits 6 Visits -1 0 1 2 3 4 5 6 7 UNS** EOS*** 7 8 Weeks -2 to 0 0 8 16 24 32 40 48 56 -- 56-60 9 10 Informed consent X 11 12 Inclusion & exclusion criteriaFor peerX review only 13 14 15 Randomization X 16 17 Medical history X http://bmjopen.bmj.com/ 18 19 Physical examination X X X X X X X X X X 20 21 22 Echocardiography X 23 24 INR X X* X* X* X* X* X* X* X* X 25 on September 26, 2021 by guest. Protected copyright. 26 Renal function X X X X X X 27 28 29 Drug dispensing X X X X X X X X 30 31 Drug collection X X X X X X X X 32 33 Outcome events X X X X X X X X X 34 35 36 Adverse events X X X X X X X X X 37 38 39 40 41 42 25 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

BMJ Open Page 28 of 55

17 http://bmjopen.bmj.com/ 18 19 20 21 22 23 24

25 on September 26, 2021 by guest. Protected copyright. 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 26 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 29 of 55 BMJ Open

1 2

27 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 55

28 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 55 BMJ Open

29 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 55

30 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 55 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Figure 1. Design for the DAVID-MS study

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 26, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 55 BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 55

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 55 BMJ Open

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 55

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 55 BMJ Open

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 55

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-038194 on 25 September 2020. Downloaded from

Page 41 of 55 BMJ Open

BMJ Open Page 42 of 55

17 groups overseeing the trial, if applicable (see http://bmjopen.bmj.com/ 18 Item 21a for data monitoring committee) 19 20 21 22 23 24

Page 43 of 55 BMJ Open

BMJ Open Page 44 of 55

Page 45 of 55 BMJ Open

BMJ Open Page 46 of 55

Page 47 of 55 BMJ Open

17 http://bmjopen.bmj.com/ 18 19 20 21 22 23 24

BMJ Open Page 48 of 55

Page 49 of 55 BMJ Open

17 http://bmjopen.bmj.com/ 18 19 20 21 22 23 24

BMJ Open Page 50 of 55

Page 51 of 55 BMJ Open

BMJ Open Page 52 of 55

Page 53 of 55 BMJ Open

BMJ Open Page 54 of 55

17 http://bmjopen.bmj.com/ 18 5. Primary Sponsor None 19 20 6. Secondary Sponsor(s) None 21 22 7. Contact for Public Queries Prof Chung-Wah SIU, MD 23 24 Cardiology Division,

Page 55 of 55 BMJ Open

BMJ Open Page 56 of 55

Page 57 of 55 BMJ Open