Transcatheter aortic valve implantation complexity score

Introduction

Transcatheter aortic valve intervention (TAVI) has revolutionised the treatment of severe aortic stenosis (AS) and is now an established therapy for patients across the full spectrum of surgical risk.1–3 As a result, TAVI has become the dominant form of aortic valve intervention, with the volume of TAVI procedures carried out exceeding all forms of surgical aortic valve replacement in the USA since 2019 and the UK since 2020.4 5 The forecast is for this expansion to continue; while the patient cohorts eligible for TAVI continue to diversify, recent studies have also highlighted the significant burden of undiagnosed severe AS, with the UK estimated to have 300 000 people who are potentially in need of treatment.6 As a consequence, the pressure on healthcare systems to deliver TAVI will only increase in the coming years and decades, placing an ever-greater emphasis on the efficiency of associated pathways and services.7

One of the main limitations to increasing TAVI capacity is the availability of cardiac catheterisation laboratory time. As such, there is an increasing incentive to maximise available slots by performing as many cases per list as is safely possible and to minimise the number of cancellations. An important way of achieving this is to ensure that patients are listed appropriately based on how long cases are forecast to take and the probability of significant procedural complications. Performing four or five uncomplicated transfemoral TAVI cases in a single day is now routine across many centres, but certain cases may be more complex and require additional time; scheduling these in an already pressurised list may lead to cancellation of patients on the day and result in wasted resources. As a result, we have developed a novel TAVI ComplEXity scoring system (TEX score) which can aid in planning lists with appropriate case selection.

The TEX score is based on a combination of patient factors (eg, body mass index (BMI), clinical instability, left ventricular ejection fraction (LVEF)) and technical factors derived from preprocedural cardiac-gated CT scanning (eg, annular angulation, ileofemoral disease). The scoring system separates patients to three levels of rising complexity. In order to validate the TEX score, we have undertaken a retrospective analysis of TAVI cases carried out at a large TAVI centre in the UK with the hypothesis that increasing TEX score correlates with increased procedural duration and reduced valve academic research consortium (VARC) 3 technical and device success.

Methods

Development of the TEX score

The TEX score was developed based on the consensus of four experienced TAVI operators, together with published evidence on factors influencing procedural outcomes. Patients are assigned a complexity level of 1 (low), 2 (intermediate) or 3 (high) based on the presence or absence of specific clinical and anatomical variables, listed in figure 1. The scoring system is hierarchical rather than points based, such that the presence of the highest-ranked factor determines a patient’s complexity level. For example, a patient who has a bicuspid aortic valve (BAV) is classified as level 2 complexity and remains at this level even if they have other level 2 features such as challenging transfemoral access. However, if the same patient was undergoing TAVI via an alternative access route, they would then be classified as level 3 complexity. The scoring is done by a consultant TAVI operator during the Heart Team meeting.

Figure 1

Request permissions

The component of the TEX score. *Critical pre-operative status is defined as any of the following options: ventricular tachycardia or ventricular fibrillation or aborted sudden death, pre-operative cardiac massage, pre-operative ventilation before anaesthetic room, pre-operative inotropes or intra-aortic balloon pump, pre-operative acute renal failure (anuria or oliguria <10 mL/hour). AR, aortic regurgitation; AV, aortic valve; BASILICA, bioprosthetic or native aortic scallop intentional laceration to prevent iatrogenic coronary artery obstruction; LVEF, left ventricular ejection fraction; NEWS, National Early Warning Score; TEX, TAVI ComplEXity; TF, trans-femoral; ViV, valve-in-valve.

Patients in level 1 are deemed to be ‘straightforward’ cases which are suitable for a list of 5 in a day with nurse-led sedation. Those in level 2 are more complex and may be better suited to a list with fewer patients, allowing more time per case. Patients in level 3 are complex and generally require anaesthetic support as well as more procedural time.

Results

A total of 1034 consecutive patients who underwent TAVI between 1 June 2021 and 31 October 2023 at Leeds Teaching Hospitals NHS Trust were included in the validation study. When the TEX score criteria were applied to this cohort, 582 (56.3%) were classified as level 1 complexity, 377 (36.5%) as level 2 complexity and 75 (7.3%) as level 3 complexity.

When assessing the total patient population, the majority (61.9%) were male and the mean age was 80.4 years (table 1). The dominant vascular access type was percutaneous transfemoral (97.1%), with percutaneous axillary the next most commonly used (2.2%). Elective surgical cutdown, whether femoral or subclavian, accounted for only 1.5% of cases. The most frequently used valve type was Medtronic Evolut Pro/Pro+ (46.1%), followed by Edwards SAPIEN 3/Ultra (34%) and Abbott Portico/Navitor (17.6%). The remaining cases were performed using Acurate Neo/Neo 2 and JenaValve Trilogy systems.

Comparing baseline characteristics between patients in the three levels of complexity, the only significant difference between the groups relates to previous cardiac surgery; a greater proportion of patients within complexity level 2 had undergone cardiac surgery than those in levels 1 or 3 (6.7% vs 20.4% vs 13.3%; p<0.05).

There were differences observed between the three groups with respect to type of transcatheter heart valve used. A lower proportion of patients in the level 3 complexity group were treated with an Edwards SAPIEN 3/Ultra compared with levels 1 or 2, and the overall difference between the groups was significant (33% vs 38.5% vs 18.7%; p<0.05). Conversely, a higher proportion of patients in the level 3 complexity group were treated with Abbott Portico/Navitor valves compared with levels 1 or 2, and once again the overall difference between groups was significant (21.1% vs 9.8% vs 29.3%; p<0.001). No significant difference was observed with respect to the use of Medtronic Evolut R/Pro/Pro+ valves across complexity groups (43.3% vs 50.1% vs 48%; p=0.11).

With respect to the primary endpoint of procedural duration, a significant difference was observed between the three groups, with duration rising with each incremental increase in complexity level (73.7 min vs 85.6 min vs 136 min; p<0.001) (table 2 and figure 2). Only a low proportion of patients in levels 2 and 3 complexity groups had a procedural duration which was lower than the mean procedural duration for level 1 patients (14.3% (54/377) and 12% (9/75), respectively).

A significant difference was noted between complexity levels with regard to the secondary endpoint of VARC-3 technical success (98.1% vs 97.6% vs 92%; p<0.05). When examining the constituent components, this difference was driven by lower rates of successful access, device delivery and retrieval (99.5% vs 99.2% vs 94.7%) and freedom from heart surgery and major vascular surgery (99.3% vs 99.2% vs 94.7%) in level 3 complexity patients. A significant difference between the three groups was observed with respect to VARC-3 device success (96.2% vs 92.3% vs 86.6%; p<0.001), but not VARC-3 early safety (88.9% vs 86.2% vs 84.0%; p=0.31) (figure 3 and table 2).

Discussion

The volume of TAVI procedures being carried out worldwide has been rising year-on-year, and that growth is projected to continue for the foreseeable future. In the UK, demand for TAVI already far outstrips capacity, resulting in long waits for treatment. In 2019, a survey of 23 UK centres highlighted a median waiting time from referral to TAVI of over 20 weeks.7 Given that the median survival of severe symptomatic AS is worse than that of most metastatic cancers, it is not surprising that long delays to treatment translate into excess mortality; a delay to TAVI of 6 months results in almost a quarter of patients (23.3%) dying on the waiting list.27

Creating new capacity for complex procedures such as TAVI in an already-stretched healthcare system is difficult, so maximising the current level of resources to deliver as many procedures as possible must be a priority for all TAVI centres. Increasing the number of cases carried out per list is an obvious way to boost procedural numbers, but to do so safely and in a way that is sustainable can be challenging. One of the key considerations is to schedule an appropriate combination of cases, based on how long each case is predicted to take and the skill mix of staff available during that list. To this end, we have developed the TEX score, a novel scoring system to enable structured listing of cases. The TEX score uses a combination of specific clinical and anatomical variables to assign patients into three levels of rising complexity. At our institution, each patient listed for TAVI is assigned a complexity level at the time of Heart Team discussion, enabling specialist valve nurses to schedule appropriate lists according to available time and available staff. For example, five level 1 cases can be scheduled for a routine TAVI list with nurse-led sedation, whereas a level 3 case, such as one requiring chimney stenting, would be scheduled to be part of a list with fewer cases and dedicated anaesthetic support.

In order to assess the validity of the TEX score, we undertook a retrospective analysis of over 1000 TAVI cases. The stratification of patients based on the TEX score correlated with procedural duration, such that each incremental increase in TEX score was associated with prolonged procedure time. The TEX score also correlates with the incidence procedural complications; significant differences were noted between groups with respect to VARC-3 technical and device success. The difference in VARC-3 technical success was of particular interest since this relates directly to immediate success of a procedure at the time when the patients leave the operating room. As such, being able to identify patients who are at risk of complications that result in this endpoint not being achieved may influence the planning of lists. From our analysis, the main factor driving the difference in VARC-3 technical success was the increased incidence of vascular complications among patients in the level 3 complexity group.

While established, pre-existing risk scores, such as the Society of Thoracic Surgeons or EuroSCORE II, have been applied to TAVI in order to predict outcomes,28 there remains a paucity of complexity scoring systems with the aim of predicting procedural duration and complication rates in order to planning appropriate scheduling. A scoring system based on CT findings, named TAVI-CT score and developed by Corcione et al, has similarities to the TEX score presented here.29 It also stratified patients into three groups, and a number of the variables which contribute to the score are shared (eg, annulus angulation, ileofemoral disease, landing zone calcification). In a prospective validation study of their scoring system, the group demonstrated that vascular complications were significantly more frequent among those in the highest TAVI-CT scoring group. However, the present paper differs in two key areas: the TEX score comprises a combination of CT and clinical factors, rather than just the former, and our study demonstrates a clear relationship between increasing complexity score and procedural duration.

While the finding that more complex procedures take longer and have a higher incidence of complications than more straightforward ones is fairly intuitive, being able to categorise patients using a simple scoring system is potentially of great utility. It can allow an ambitious but achievable number of cases to be scheduled, while potentially reducing the risk of cancellations due to lists over-running.

Conclusion

The TEX score is a simple tool which allows stratification of patients into three levels of complexity. Increasing complexity levels correlate with increasing procedural duration and reduced VARC-3 technical and device success. This is potentially useful for scheduling patients onto appropriate lists to allow effective utilisation of cardiac catheterisation laboratory resources and ultimately increased procedural numbers.