Do we need disease-specific, generic single-brand thoracic stent-graft registries?

Grassi and Weaver [1] recently reported the midterm outcomes of thoracic endovascular aortic repair (TEVAR) of descending thoracic aortic aneurysms (DTAAs) from the Gore Global Registry for Endovascular Aortic Treatment (GREAT). From a total of 310 patients, the Gore thoracic aortic device was associated with a low rate of device-related reinterventions and aorta-related mortality for up to 5 years [1]. The GREAT registry is a multicentre-sponsored prospective observational cohort registry to collect data on the incidence of device-related events from a ‘real-world perspective’. Nowadays TEVAR has been established as the standard of care for different thoracic aortic diseases [2, 3]. Notably, guidelines with a high quality of evidence (level A) and a strong recommendation level (grade 1) for TEVAR as the preferred approach for DTAAs have been waived in favour of non-randomized studies and controlled prospective trials [3, 4]. Consequently, industry-sponsored studies or registries will play only a minor role in building adjunctive evidence to support TEVAR as useful and effective.

Global registries have many advantages, but their reports must be analysed with caution. Registries allow the collection of data from large cohorts at lower costs, can be easily set up compared to controlled studies, and could potentially provide evidence of the effectiveness of treatments. Moreover, including data from many worldwide centres might provide real-world insights into the morbidity and mortality rates, thereby eliminating the high-volume centre bias present in a prospective controlled trial. Unfortunately, registries present several limitations that should be addressed when critically analysing reports based on these data.

First, patients are seen at different centres and in different countries. The assessment or treatment criteria may not be uniform among the participating centres, which could potentially introduce selection bias. In addition, local conditions for data collection and submission vary from centre to centre, and often, there is a lack of data monitoring and validation. Second, patients are not strictly monitored like they are in randomized controlled studies, and the absence of a follow-up scheme and reintervention protocols in the registry may underestimate the rate of late adverse events. When assessing the mid or long-term outcomes of a procedure, complete and long-term follow-up is crucial but remains a challenge in global registries. Indeed, Grassi and Weaver [1] reported that only a few patients reached the 5-year follow-up, confirming these limitations. Lastly, the statistical analysis of registry data focuses on generic and early end points. The data are managed by the company sponsoring the registry whose goal is to study the performance of the device rather than to provide a targeted subanalysis.

Although these limitations might reduce the scientific interest of such generic analyses, the outcomes analysis of new devices introduced in daily practice is of paramount importance. On the one hand, innovative endovascular devices might increase the clinical success rate of treating complex thoracic diseases by, for example, improving the profile of the sheaths or increasing the conformability in the distal arch, thereby reducing the endoleak rates. On the other hand, the uncontrolled use of a new endovascular armamentarium must always be analysed with caution, and the benefits and the risks arising from new devices must be carefully weighed, especially concerning their long-term durability. We recently witnessed the massive recall of a thoracic stent graft due to stent fractures and endoleak [5]. Indeed, new European Union medical device regulations with more strict protocols went into effect in May 2021, mainly for manufacturers, to assure the quality, performance and safety of implanted medical devices. The new regulations require companies to have quality management systems in place, including a ‘clinical evaluation’ for each device, monitoring and measurement of data analysis and outcomes and processes for reporting incidents and corrective actions [6].

Besides the above-described global registry concerns, other issues are still present in many aspects of the management of thoracic aortic diseases and need to be elucidated. For example, according to the guidelines of the European Society for Vascular Surgery, TEVAR for uncomplicated acute type B aortic dissection or for patients with underlying genetic diseases, as well as the best method of repair of ruptured DTAAs, still needs to be clarified [7, 8]. Grassi and Weaver [1] failed to report disease-specific comparisons and multivariate analyses due to the few adverse events determined by the good TEVAR outcomes. Multiple subgroup assessments after TEVAR are crucial in the overall context of the treatment of thoracic aortic diseases, and we should acknowledge that data analysis in the management of different thoracic aortic diseases is still needed to promote strong, clear evidence. In the future, we hope to see the 10-year follow-up outcomes analysis from the GREAT registry as well as from the other studies, with disease-related subgroup analyses and long-term assessment, which may provide more reliable data and robust evidence for TEVAR.

Conflict of interest: none declared.

REFERENCES