Rivaroxaban or Aspirin for Biological Aortic Prosthesis

Background

The study is a comparison of aspirin and rivaroxaban for the postoperative treatment after biological aortic valve replacement.

Scientific Background Guidelines have over the years recommended anticoagulation or aspirin after surgical replacement of the aortic valve with a prosthesis. There are no proper randomized trials. Rivaroxaban is an anticoagulant approved for several indications. The investigators are conducting a study to compare Aspirin versus Rivaroxaban for prevention of thrombosis during 6 months after aortic valve replacement with a biological prosthesis.

As a substudy CT scans for a subgroup of patients to detect partially immobilized valves due to thrombosis is performed.

Trial drugs The trial drugs are 100 mg Aspirin daily or 10 mg Rivaroxaban daily. Patients are provided with either 180 tablets of 10 mg rivaroxaban or 180 tablets of 100 mg aspirin.

Methodology

This study is planned as a multicenter, randomized, open label, parallel group study. Patients are allocated to aspirin or rivaroxaban to be treated for 6 months. Treatment after this period is not trial driven and left to the discretion of the investigator. As indicated below the study is planned to include 1000 patients. The steering committee may review recruitment rate and overall event rate (blinded by treatment) and may increase the sample size to up to 2000 patients. Another important aspect of this study is that follow-up is mainly by registers. This is a group of patients who will in all cases receive treatment similar to that tested in the trial and where there is systematic follow-up in cardiology clinics. Therefore systematic follow-up only directed by the trial is kept to a minimum and most information during follow-up is obtained from registers.

Randomisation For all patients that have provided informed consent a web-based case record form is completed with demographic data, data regarding the heart disease, comorbidity, medication and time of study start.

Randomisation is performed by a computerized algorithm with a varying block size and stratification by operating center. The specific treatment of each patient is open and provided by the computer when the patients is randomised.

Patients are considered enrolled in the study when they have signed the informed consent. All patients that have signed the consent will be accounted for when the study is reported. In terms of the primary intention to treat analysis patients are part of this analysis when they take the first tablet of either aspirin or rivaroxaban. An analysis including all patients that have provided informed consent will be a sensitivity analysis. The earliest time of study start is postoperatively after removal of temporary pacemaker electrodes. This is anticipated to be a few days after the operation. Study can start up to 2 weeks later when anticoagulation in the intermediate period has only been low molecular weight heparin and/or aspirin.

Follow-up Study staff contacts patients after 3 and 6 months by telephone to ensure compliance and to obtain reports of events including adverse events.

Study end The study ends after 6 months. The final contact to study staff is by telephone. There are no guidelines indicating that further anticoagulation should be necessary for these patients as directed by the valve operation and there is no recommended further treatment of the patients or further follow-up by direct control. Follow-up in registers will continue after study end.

There is a preplanned analysis of outcome by treatment allocation at 2 years.

Statistical analyses

The primary comparison is a log-rank test of time to selected outcome. Supportive analyses will use Cox proportional hazard models. For an analysis of superiority of either drug a 2 sided p-value of 0.05 will be used. The conclusion of superiority is based on the main endpoint.

A statistical analysis plan will be prepared prior to study closure. The main analyses will be made according to the intention to treat principle where all patients that have taken at least one tablet will be examined in the main analysis.

In addition analyses will be made where patients are censored when they stop study medication.

Sample size:

Access to nationwide registered were employed. A total of 2345 patients treated with aspirin and 4509 treated with warfarin after operation were identified. Analysis was performed between 30 days and 180 days to avoid conditioning on the future. It should be noted that this procedure most likely underestimates the power of the study. The 30-180 day primary event rate in the aspirin group was 7.5% and the event rate in the warfarin group was 3.2%. Based on these observations the CPOWER function of the R-HMisc package was used to calculate a power of 81% based on 500 simulations.

Safety Management A Data Safety and Monitoring Board with statistical, cardiological and surgical expertise will follow the study. Formal interim analyses are not planned. This committee is composed of members otherwise independent of the study. They can recommend the steering committee to discontinue the study, but it is the only body which during the course of the study received unblinded results grouped by treatment.

As for the individual patient the investigator may discontinue study treatment if bleeding is observed or otherwise indicated. There are no trial specific rules on how to handle side effects during the trial.

CT substudy A CT substudy of 580 patients will be conducted to investigate the effect of postprocedural medical regime on leaflet motion in the aortic bioprosthesis. These patients follow the normal course of the trial and further have a CT scan of the aortic valve performed after 3 and 12 months. Patients are consecutively invited to this substudy until the required sample size is met.

Primary end-point • The proportion of patients with at least one prosthetic leaflet with >50% motion reduction as assessed by cardiac 4DCT-scan (total N = 580).

Secondary end-points • The proportion of prosthetic leaflets with >50% motion reduction as assessed by cardiac 4DCTscan (total N = 1740). • The proportion of patients with at least one prosthetic leaflet with thickening as assessed by cardiac 4DCT-scan (total N = 580). • The proportion of prosthetic leaflets with thickening as assessed by cardiac 4DCT-scan (total N = 1740). • Aortic transvalvular mean pressure gradient and effective orifice area (cm2) as determined by transthoracic echocardiography. • Functional NYHA class. • Death, first thromboembolic event (DTE), and safety endpoints will be assessed in the main study and analysed in the CT substudy with regards to occurrence of the leaflet abnormalities.

Definitions • Leaflet thickening: Hypoattentuating leaflet thickening or focal hypoattenuating abnormality attached to the prosthetic leaflet or diffuse thickening of the prosthetic valve leaflet identifiable on at least two reconstructed planes (typically double-oblique axial and multiplanar reformatted reconstructions). • Reduced motion: Reduced systolic leaflet excursion is classified as: (I) normal, (II) mildly reduced (<50%), (III) moderate to severely reduced (>50%), and (IV) immobile. Reduced systolic leaflet excursion is considered significant when it is > 50% or immobile. Quantitative assessment of leaflet motion is performed with a blood pool inversion volume rendered cine reconstruction throughout the cardiac cycle evaluating the bioprosthetic leaflets.

The primary endpoint of reduced aortic bioprosthetic leaflet motion is expected to occur in approximately 10% of patients.(15) Equal number of patients is warranted in both treatment groups. A 65% reduction during rivaroxaban of primary endpoint requires inclusion of 580 patients in order to demonstrate superiority. Taking into account an estimated loss at follow-up of 18% of patients, the total sample size (N) needed to demonstrate superiority would be 580 randomized patients. The primary endpoint will be analysed using an exact Fisher test or a Χ2 test, as required. Supporting analyses will include logistic regression with inclusion of covariates.

Sample size calculation (Power 0.8, Alpha 0.05)

• Probability of event in Group A: 0.10 • Probability of event in Group B: 0.035 • Proportion of sample in Group A: 0.50 • Evaluable subjects needed in Group A: 238 • Drop-out: 0.18
• N = 580

Estimation of dropout rate • A 10% loss at follow-up or death at 3 months is expected in this elderly patient population. • In approx. 8% of cases, an inconclusive cardiac 4DCT can be expected.

Coordinating centre Rigshospitalet - University Hospital, Copenhagen, Denmark