Table 1

Advice for patients who are suitable for SAVR or TAVI*

Characterisation of patient populations

The following inclusion criteria have been established for the trial: patients must have severe, calcific, symptomatic AVS, with or without the need for concomitant coronary artery bypass graft (CABG) surgery or percutaneous coronary intervention (PCI), who are at low operative risk for SAVR. Furthermore, patients must have undergone either of the following surgical approaches: an open surgical approach with the use of a conventional stented xenograft bioprosthesis or a sutureless prosthesis (Perceval, LivaNova plc, London, United Kingdom); or transcatheter aortic valve implantation with either a balloon or a self-expanded transcatheter heart valve (THV). Participants of any gender, race or ethnicity are eligible for inclusion in the study.

Endpoint definitions and measurement

Study procedure

Therapeutic interventions

Valve implantation should take place before 14 and 21 days after enrolment for prospective allocation and no later than 30 days after informed consent. The date of valve implantation will be considered as day 0. The preliminary encounter (day 0) is designated for the scheduling of all subsequent encounters and the calculation of visit windows. Patients who undergo SAVR or TAVI will remain enrolled in the study and will complete it through year 10, in accordance with the visits and events delineated in the study procedure and schedule of procedures. On the initial day (day 0) of the VI procedure, a comprehensive evaluation of the patient’s cardiovascular system will be conducted. This evaluation will encompass the administration of medications designed to regulate cardiovascular function and antithrombotic/anticoagulant therapies. It will also include an assessment of potential adverse events (AEs). The evaluation will be further supplemented by a TTE or TEE as well as a supra-aortic angiogram or TEE, as deemed necessary by the attending medical team. It is recommended that patients participating in the study should receive prophylactic treatment against endocarditis in line with recommendations issued by the American Heart Association.35 36

As illustrated in table 3, the recommended anticoagulation/antithrombotic regimen is presented.

Standard AVR procedure

Replacement of a faulty aortic valve typically involves the insertion of a synthetic graft prosthesis, a conventional stented xenograft bioprosthesis or a rapid-deployment sutureless prosthesis. Concomitant surgical interventions, such as CABG, the treatment of atrial fibrillation, septal myotomy and aortic root enlargement, are deemed permissible in such cases. Further details on the procedure are consulted in online supplemental material S1.

For patients undergoing SAVR, the standard of care as outlined by the institution dictates the usage of a bioprosthetic surgical valve and associated components that are commercially available.

The Perceval sutureless prosthesis (manufactured by LivaNova plc, a United Kingdom-based company) will be implanted in patients diagnosed with severe symptomatic AVS (online supplemental figure 2). In order to minimise the impact of selection bias, a CT scan was performed during the enrolment phase prior to the implantation of the SAVI. This scan confirmed the eligibility for the current sutureless valve implantation, the suitability for the proposed surgical access (full sternotomy or ministernotomy) and the decision regarding an isolated or concomitant procedure. The use of a right anterior minithoracotomy was precluded due to the variable experience of the centres with this procedure as well as its unsuitability for the purpose of serving as a comparator to the standard valve. Further elucidation on the sutureless valve and implantation procedure may be found in the supplementary material (online supplemental material S1).

Transcatheter aortic valve implantation

The utilisation of both the balloon and the self-expanding THV is contingent on the Edwards SAPIEN Transcatheter Heart Valve (Edwards Lifesciences LLC One Edwards Way Irvine, CA 92614 USA)6 7 18–23 and the CoreValve (US CoreValve Clinical Investigators trial, Medtronic) being the preferred choice.8 23 24 The SAPIEN 3 THV (PARTNER consortium—Placement of AoRTic TraNscathetER Valve Trial, Edwards SAPIEN Transcatheter Heart Valve) is a catheter-delivered heart valve that combines a balloon-expandable stent and bioprosthetic valve technology. Transcatheter delivery of the study valve is facilitated via transfemoral access. The device’s composition is delineated below: first, a radiopaque, cobalt-chromium alloy balloon-expandable frame, second, a trileaflet bovine pericardial tissue valve, third, a polyethylene terephthalate (PET) internal fabric skirt; and fourth, a PET external sealing ring. The utilisation of self-expanding, supra-annular bio-protheses, such as the CoreValve, Evolut R or Evolut PRO models is to be employed. The transcatheter bioprosthesis consists of a self-expanding nitinol frame and a porcine trileaflet pericardial valve. For an in-depth exploration of the TAVI procedure, including implantation, please refer to the online supplemental material S1, online supplemental figures 3–5 provided in the online resource.

Box 2

Systems

Cardiopulmonary

Clinical lab testing

Neurological evaluations

Functional evaluations

Box 3

Serious adverse event (SAE)

An adverse event is deemed to be serious if the events are associated with

• Fetal distress, fetal death, congenital abnormality or birth defect.

• Medically significant incident.

Serious medical events not meeting the above criteria may still be SAEs if they endanger the patient and require immediate medical or surgical intervention to prevent one of the above outcomes.

Pre-existing medical conditions or symptoms reported before enrolment will not be recorded as an AE. Record an AE if the pre-existing condition or symptoms worsen due to the device, or if study symptoms are due to the device or trial-related procedure. Do not record death as an AE, but only as a consequence of another specific AE.

Anticipated adverse events (AEs) are defined as such events which have been identified as potentially occurring in connection with the investigational device or implant procedure.

An unanticipated adverse device effect (UADE) is defined as any serious adverse effect on health or safety, or any life-threatening issue or fatality, caused by, or associated with, a device, if that effect, problem or death was not previously identified in the investigational plan or application (including a subsidiary plan or additional application) or any other unanticipated serious issue related to the rights, safety or welfare of subjects.

It is imperative that all UADEs are reported to CEC without delay. Completion of the AE Form of the CEC is also mandatory for all UADEs. In addition, the Investigator is obliged to inform his/her EC/IRB of all UADEs occurring at his/her site no later than 10 days after the Investigator first becomes aware of the effect (and any additional information as required by EC/IRB or local regulations).

All AEs associated with UADEs are subjected to close monitoring until either resolution or a stable clinical endpoint is achieved. Essential treatments and outcomes must be documented.

The absence of reliable evidence in extant guidelines was the motivation for this study

Since Cribier’s first TAVI procedure in 2002,40 over 1.500.000 patients have received TAVI worldwide.41 The development of international guidelines recommending TAVI over conventional surgery for the treatment of severe AVS has been supported by robust research findings.6–8 18–25 A comprehensive and rigorous programme of studies, involving multiple randomised, multicentre trials, has provided substantial evidence in support of this recommendation. These recommendations are supported by Class of Recommendation (COR) I, LOE A, which is the highest level of recommendation available and implies the substantial safety and efficacy of the procedure.2 3 By contrast, with the number of valves implanted set to reach 75 000 by 2022, the figure of patients who have undergone SAVI since the first Percevalve sutureless valve implantation in 2007 is considerably lower.42 Additionally, a notable discrepancy exists in the robustness of results between the smaller number of SAVI and TAVI implants, with the former supported by a limited number of multicentre RCTs.43–45 The implications of these observations are significant for the safety and efficacy recommendations outlined in international guidelines and supported by COR 1, LOE A.2 3

In meta-analyses of observational data, sutureless valves appear to demonstrate certain benefits in comparison with both conventional SAVR and TAVR. Similar to TAVR, the midterm outcomes of sutureless valves are deemed satisfactory; however, the long-term results remain pending.46 47

The primary benefit of sutureless valve technology as perceived by researchers is the reduction in ischaemic surgical time when compared with conventional sutured valves.48 However, there is an absence of compelling data to suggest that the decrease in aortic cross-clamp time, facilitated by the use of the sutureless valve, may result in improvements in morbidity or mortality.46 47 49–53 A recent study has revealed that cross-clamp times were found to be significantly reduced in the SAVI group when compared with the full sternotomy SAVR group. However, no significant differences were observed in terms of cumulative CPB time or clinical outcomes.45 These findings were derived from the CADENCE-MIS trial,45 a randomised study that compared minimally invasive SAVI with rapid deployment valves with full sternotomy SAVR. Nonetheless, an association between prolonged cross-clamp or CPB times and an increased risk of postoperative complications, including renal failure, respiratory failure, low-output syndrome, postoperative atrial fibrillation, higher transfusion requirements, longer postoperative hospital stays and trends in mortality, has been suggested by several retrospective observational studies.54–56⁻ A systematic review and meta-analysis of randomised and propensity-matched comparative studies was conducted with the objective of investigating the impact of diminished operative times on clinical outcomes. The analysis demonstrated that equivalent 1-year survival outcomes were observed; however, diminished postoperative complications, such as atrial fibrillation and blood product transfusions, were observed in the SAVI group. In contrast, augmented PMK implantation was noted.57

This study explores the contemporary interest among surgeons in the use of the SAVI system. It is challenging to ascertain whether the inclination towards SAVI is predominantly influenced by the simplicity of implantation and the surgeon’s perception of diminished operative time in high-risk patients, as opposed to its tangible effect on clinical outcomes. This inquiry necessitates the execution of ad hoc studies. However, it is plausible to hypothesise that SAVI could offer a potential advantage in terms of postoperative complications for selected high-risk patients requiring multiple or extensive procedures.

Sample size calculations

The sample size for the trial has been calculated on the basis of attaining a minimum of 85% power in order to pass the 1, 5 and 10-year safety and effectiveness endpoint. The event rate estimates for the primary safety and effectiveness endpoint have been derived using data from prior studies. Given that the current study population is a lower risk cohort and in order to account for both procedural refinement over time in both groups and changes in definitions for components of the endpoint, the rates have been adjusted.

The sample size is based on the composite primary endpoint, which assumes an event rate of 16.6% in the SAVR arm and 14.6% in the TAVR cohort. An enrolled sample size of 1.757 patients with 10-year data would produce 85% power for passing the endpoint. The calculation of this sample size is predicated on the utilisation of a one-sided Score test (Farrington & Manning), with a significance level of alpha=0.025, and incorporating the stipulated non-inferiority margin of a relative 35%.

The primary sample size estimation was derived through a pure frequency analysis; however, the endpoint analysis will use the Kaplan-Meier estimates. Given the equivalence of the two methodologies in the absence of censoring and the anticipated sufficiency of uncensored patients, the sample size is expected to be adequate. However, given the highly uncertain feasibility assumptions associated with this previously unstudied population, an actual sample size of 2.020 has been determined, representing a 15% increase over the minimum size deemed statistically justified. This augmented sample size is intended to not only meet the requirements of the study but also to provide additional flexibility in case of contingencies, such as withdrawals or losses to follow-up, which are inherent in clinical trials. Furthermore, this selected size is expected to enhance the precision of analysis for various unpowered secondary endpoints.

Analysis populations

The ITT population constitutes all randomised patients, whereas the As Treated (AT) population is a subset of the ITT population consisting of all patients for whom the index procedure is initiated, irrespective of whether the index procedure is completed. In instances where multiple procedures are undertaken, the final procedure that involved the deployment of the study valve will be designated as the index procedure. This procedure will then be utilised to determine the AT trial allocation, and the date of the index procedure will be employed in the determination of all subsequent follow-up visits and associated assessments.

The VI population constitutes the subset of the AT population consisting of all patients who receive and retain the intended valve during the index procedure. Patients who receive a valve in VI will thus be part of the VI population. Furthermore, patients who are converted from TAVI to SAVR during the procedure will also be part of the VI population. The AT population is to be used as the primary population for trial endpoint analysis. The VI population will be used for analysis of echocardiographic data and related endpoints, while selected sensitivity analyses will be performed using the ITT population.

Endpoint analysis

The composite endpoints to be evaluated include all-cause mortality, total stroke and rehospitalisation rates at 1, 5 and 10 years following procedure completion. These endpoints will undergo a non-inferiority analysis with a relative non-inferiority margin set at 35%, as outlined in the study protocol. The components of the composite endpoint will be methodically analysed by the CEC. Patients will be classified as having experienced the endpoint once any component of that endpoint is documented. In the absence of such an event, the patient will be considered to be free of the endpoint and censored for the subsequent analysis, which will be based on the last known date on which the patient was alive and free of events.

In conducting the event rate difference endpoint test, the non-inferiority margin of a relative 35% will be utilised. A one-sided non-inferiority test at an α-level of 0.025 will be performed, which involves the computation of the two-sided 95% confidence limit for the event rate ratio (TAVI/SAVR). It is imperative to note that the acceptance criterion necessitates the upper confidence limit to be no greater than 1.35. Should the non-inferiority analysis demonstrate a satisfactory outcome, the subsequent superiority analysis will be conducted. The type I error rate for this analysis is protected by the non-inferiority analysis, thereby precluding the necessity for an alpha adjustment.

The secondary endpoints will be categorised into two groups. For the key secondary endpoints, testing for superiority will be conducted in a prespecified hierarchical order, with the use of a gatekeeping method to control for multiple comparisons. P values will be presented alongside claims of significance. Conversely, for other secondary endpoints, analyses will be conducted without the application of a correction for multiple comparisons. Consequently, HRs and 95% CIs will be presented without p values or claims of significance. It is important to note that inferences derived from these 95% CIs may lack reproducibility.

The comparison of continuous variables, which will be presented as means with SD or medians with interquartile ranges, will be conducted using Student’s t-test or the Wilcoxon rank-sum test. Categorical and ordinal variables, which are presented as proportions, will be compared using Fisher’s exact test or the Wilcoxon rank-sum test. Following the implementation of a baseline, continuous variables will be analysed through the utilisation of an analysis of variance, with adjustment for the baseline measurement. Time-to-event analyses will be performed using Kaplan-Meier estimates and then compared using a log-rank test. Median survival time and area under the Kaplan-Meier curve were used to estimate life expectancy and valve durability. Risk factors were identified using univariate and multivariate Cox proportional hazards regression. Echocardiographic analyses will be conducted on the VI population, comprising patients in whom the intended valve was inserted. Statistical analyses will be conducted utilising R software (R Foundation for Statistical Computing, Vienna, Austria).