Article Text
PDF
Abstract
In elderly (75 years or older) patients living in Latin America with severe symptomatic aortic stenosis candidates for transfemoral approach, the panel suggests the use of transcatheter aortic valve implant (TAVI) over surgical aortic valve replacement (SAVR). This is a conditional recommendation, based on moderate certainty in the evidence (⨁⨁⨁Ο).
This recommendation does not apply to patients in which there is a standard of care, like TAVI for patients at very high risk for cardiac surgery or inoperable patients, or SAVR for non-elderly patients (eg, under 65 years old) at low risk for cardiac surgery. The suggested age threshold of 75 years old is based on judgement of limited available literature and should be used as a guide rather than a determinant threshold.
The conditional nature of this recommendation means that the majority of patients in this situation would want a transfemoral TAVI over SAVR, but some may prefer SAVR. For clinicians, this means that they must be familiar with the evidence supporting this recommendation and help each patient to arrive at a management decision integrating a multidisciplinary team discussion (Heart Team), patient’s values and preferences through shared decision-making, and available resources. Policymakers will require substantial debate and the involvement of various stakeholders to implement this recommendation.
- transcatheter aortic valve replacement
- aortic valve stenosis
- heart valve prosthesis implantation
Statistics from Altmetric.com
The following clinical and technical aspects should be considered by the Heart Team, which apply across the whole age spectrum.
Factors favouring surgical aortic valve replacement (SAVR) over transcatheter aortic valve replacement (TAVI):
Unfavourable anatomy for TAVI.
Unsuitable for transfemoral access.
Presence of concomitant valvular disease with surgical indication.
Multivessel coronary artery disease.
Patients placing a higher value on bioprosthetic valve durability or pacemaker risk.
Factors favouring TAVI over SAVR:
Older age.
High or moderate risk for SAVR.
Prior sternotomy or SAVR.
Frailty.
Patients placing a higher value on lower procedural morbidity and rapid recovery.
Rationale
Nearly 20 years have elapsed since the first transcatheter aortic valve implant (TAVI).1 In the subsequent years, this technology has become increasingly popular, supported by multiple well-designed clinical randomised controlled trials.2 This transcatheter-based intervention allows for a less-invasive approach leading to faster recovery but has increased upfront procedural costs.3 Practising physicians need guidance for clinical practice, especially in what has become a rapidly evolving field.
Most clinical practice guidelines providing recommendations on TAVI for patients with severe aortic stenosis were developed in high-income settings in which healthcare systems, resource availability, and clinical practice differ substantially compared with Latin America.4 Although we identified documents providing guidance for TAVI implementation in Latin America, they have severe methodological limitations which lead to erratic conclusions.5 For instance, a 2019 technology assessment in Argentina concluded that TAVI was not superior or even inferior to surgical aortic valve replacement (SAVR), limiting the role of TAVI to inoperable patients.6 This is markedly inconsistent with the evidence and other clinical guidelines from the rest of the world.2 4 7
There is a need for rigorously developed transparent and sensitive guidance on severe aortic stenosis management to help physicians, patients and other stakeholders dealing whit this condition in Latin America. The goal of this document is to develop a high-quality and transparent guideline to help physicians and other stakeholders concerning the use of TAVI versus SAVR in Latin America. We developed such a document following the Grading of Recommendations Assessment, Development and Evaluation (GRADE) method, a trustworthy, standardised and transparent approach for the formulation of recommendations.
Guideline methods
Setting
This guideline was conceived to be applied specifically in people living in Latin America. A total of 650 million people live in Latin America and the Caribbean. According to the world bank, the majority live in upper middle-income countries (583 million), while the rest belong to lower middle-income (40 million) or high-income (27 million) countries. The adoption of novel high-cost technologies can be challenging in developing regions mainly because of limited resource availability, technology and appropriately trained personnel.
Scope
This guideline focuses on the role of TAVI as compared with SAVR in treating patients with severe symptomatic aortic stenosis. For this purpose, we chose a scenario that represents a common clinical situation in which both alternatives are considered reasonable: elderly patients with severe symptomatic aortic stenosis eligible for transfemoral TAVI. This recommendation does not apply to patients in which there is a standard of care, like TAVI for patients at very high risk for cardiac surgery or inoperable patients, or SAVR for non-elderly patients (eg, under 65 years old) at low-risk for cardiac surgery.3 We also do not present recommendations on pure aortic regurgitation or the use of TAVI prosthesis for the treatment of other heart valves.
Audience
The primary audience of this guideline are treating physicians or Heart Teams involved in the care of patients with symptomatic severe aortic stenosis. Secondary audiences include patients and other key stakeholders such as patient organisations, hospital administrators, government, policymakers and healthcare insurance agencies (whether public or private). Although the guideline is meant to be used in Latin America, due to its thorough and transparent development process it can also be adapted to other regions, especially those with a similar degree of constraints.
Guideline team and roles
Different groups with specific roles participated in the guideline development process, including a guideline panel, a methodological team, a systematic review team, a subgroup of interest team and a writing team (see the online supplemental material for details). Four methodologists with extensive experience implementing the GRADE approach participated in the guideline (PL, AI, MAR, IJ). Interventional cardiologists and cardiac surgeons were balanced in the methodology team and guideline panel.
Supplemental material
Key societies from Latin America were integrated from the beginning and endorse the guideline, including the Sociedad Latino Americana de Cardiología Intervencionista (SOLACI) and the Sociedad Interamericana de Cardiología. The panel members were selected by representatives of each society. Two patients, one post-TAVI and another post-SAVR, without major postoperative complications were included in the panel. The addition of patients was considered particularly relevant, given the important differences between the two procedures and the lack of regional information on patients’ values and preferences.
Conflict of interest management and role of the sponsor
All guideline participants completed the International Committee of Medical Journal Editors conflict-of-interest form summarised in the online supplemental material, along with other relevant information. Participants with relevant direct financial conflict of interests, like having equity or being employed in any for-profit company (that develops, produces, markets or distributes drugs, devices, services or therapies used to diagnose, treat, monitor, manage or alleviate health conditions), holding patents or royalties were not allowed to participate in any group or process of the guideline. This guideline was funded by SOLACI, which covered costs related to the systematic review and data abstraction. Once the guideline proposal was reviewed and agreed by SOLACI for later endorsement, their board of directors was not involved in any other process of the guideline. After the guideline was finished, a quality assurance review by SOLACI was undertaken with no major modifications to the manuscript.
Involvement of other societies
This clinical guideline was conceived between a group of clinical and methods experts, who drafted a proposal which was later presented to societies for future endorsement. The content experts for guideline development included interventional cardiologists, cardiac surgeons, and heart valve disease specialists, who identified both SOLACI and SIAC as key societies. SOLACI agreed to cover costs related to systematic review and data abstraction. At the initiation of this guideline development process (in late 2019) we consulted several cardiac surgeons, including those included in the panel, who could not identify a society that represented Latin American cardiac surgeons as a whole, organised annual academic events, had their own journal, or other typical initiatives of regional societies of physicians. Therefore, cardiac surgeons from different countries were individually approached to participate in the panel.
Guideline process
We used the GRADE approach for summarising and assessing the certainty of the evidence and moving from evidence to recommendations and followed the Reporting Items for practice Guidelines in HealThcare (RIGHT) statement for reporting.8 9 To facilitate completion of assigned tasks, explicit instructions and reasonable time frames were provided to the panel members.
The final version of the guideline underwent an external review process. Four external people reviewed the guideline, including a methodological expert in guideline development with the GRADE approach, and three end-users, physicians living in Latin America: a cardiologist, a cardiac surgeon and an interventional cardiologist. Finally, the guideline underwent a quality assurance review by SOLACI representatives which approved the content of the guideline.
Guideline question
As the scope of the guideline was very specific, there was no need for question prioritisation. The methodological team agreed on the population and interventions to be considered and was confirmed with the panel before starting the review. We decided to approach the question by considering a frequent scenario that face clinicians in which, a priori, both TAVI and SAVR are acceptable alternatives. In the majority of cases, this scenario entails elderly patients. The methodological team considered that an objective age threshold is needed to define a patient as ‘elderly’ to enhance interpretation and implementation of the guideline. A decision was reached between the methodological team, topic experts and guideline panel to formulate the recommendation for elderly patients defined as 75 years or older (details expanded in the online supplemental material).
One of the most relevant decisions was not to stratify the recommendation according to surgical risk. The main reason for not pursuing such an approach was that available evidence does not suggest that the relative effectiveness of TAVI over SAVR is different across the surgical risk spectrum.2 This was also supported by our metaregression analysis showing no statistically significant association between predicted perioperative mortality and relative effectiveness of TAVI (figure 1). Although we pooled results from studies including patients with variable surgical risks, for mortality and cardiovascular mortality, we calculated absolute effects considering differential baseline risks.
Predicted 30-day mortality versus relative effectiveness of transcatheter aortic valve implant metaregression. Metaregression of Society of Thoracic Surgeons score and HR for 30-day mortality observed in each of the studies. The bubble size is proportional to the weight of each study in the fixed-effects meta-analysis.
Another relevant decision was to focus on the transfemoral approach. The transapical approach has been associated with worse outcomes compared with SAVR and was replaced with other alternatives not properly represented in randomised trials like axillary, carotid or caval.2 Other details about the guideline question process are presented in the online supplemental material.
Outcomes were rated as critical, important or not important for decision-making by the panel following the GRADE approach. Although the guideline prioritised patient-important outcomes, we decided to include paravalvular leak and patient–prosthesis mismatch as surrogates of other clinical outcomes, such as hospital admissions, reinterventions and death, given the lack of availability of long-term or very long-term data from randomised trials. Further discussion about clinical outcome selection and surrogate outcomes is detailed in the online supplemental material.
Systematic review and meta-analysis
Eligible primary studies were randomised trials comparing TAVI through the transfemoral approach with SAVR in severe symptomatic aortic stenosis. The methodological team identified a systematic review in which a comprehensive search was performed in March 2019, and a decision was made to update that search.2 A new comprehensive search was executed on Ovid MEDLINE, Embase and CENTRAL in January 2020 (see online supplemental material for an example of search strategy) and was limited to studies published from 2019. Three methodologists (PL, AI and MAR) developed the search, screened titles/abstracts and full-text in duplicate. In March 2020, UK TAVI and PARTNER 3 (2 years) results were presented at a conference and were added to the systematic review.10 11 Therefore, a total of eight randomised trials using primarily the transfemoral approach were included in the review and meta-analysis.3 10–16
Outcomes were extracted as per the Valve Academic Research Consortium 2 definitions, when available.17 The risk of bias was assessed using the risk of bias tool 2 from the Cochrane Collaboration (online supplemental table S1 in the online supplemental material).18 The certainty of the body of evidence was evaluated using the GRADE approach by the methodological team for all outcomes.8 The results critical and important outcomes were presented to the panel within a summary-of-findings table including absolute and relative risk (RR) with 95% CIs, number of participants and studies, the certainty of evidence, and interpretation.19 More details about the systematic review and meta-analysis are presented in the online supplemental material.
Evidence to decision framework
The panel discussed all relevant domains pertaining to decision-making for the question assessed following GRADE guidance.20 21 We completed an evidence the decision framework using GRADE GDT software to capture discussions and judgments detailed in the online supplemental material.22 Although we did not specifically address cost-effectiveness and impact on equity, these discussions were integrated into the resource use and acceptability domains.
For costs and resource allocation, the panel discussed the potential economic implications of implementing TAVI as opposed to SAVR considering information on local direct costs of both interventions. We did not pursue a formal economic evaluation, given the heterogeneity of health systems across Latin America, which is substantially different not only between but also within countries, which would have limited applicability. To inform the panel with updated TAVI prosthesis direct costs, representatives of industries commercialising TAVI prosthesis were approached to provide a cost range across Latin America. Given the large inconsistency in SAVR costs in this region and based on information from developed regions, we assumed that upfront TAVI costs will be substantially higher than upfront SAVR costs.23
Consensus and voting
Guideline panellists (n=13) were the only guideline members with voting rights. The experts debated on every domain of the evidence-to-decision framework, aiming for consensus. In cases when consensus did not emerge by discussion, definitive judgements were made by voting based on a simple majority (>50%). The meeting chairs (PL, AI and MAR) managed disagreements and recorded minutes or relevant aspects of the discussion, which are available in the evidence-to-decision network in the online supplemental material. Those panel members who did not attend the online meeting to debate these steps were asked to watch the full video recording and provide feedback and approval of the content, or to express disagreement if necessary.
Once panel judgements for every individual domain of the evidence-to-decision framework was defined, panellists were asked to review a short (10 min) training video on how to vote for the final recommendation. Every panellist voted recommendation’s direction and strength independently but not anonymously. Voting included three aspects: in favour or against TAVI in elderly patients with severe aortic stenosis, its strength (conditional or strong) and the main reason for their decisions. To support a strong recommendation, the methodological team pre-established that 80% or more of panel votes should be strong; otherwise, the recommendation would be defined as conditional (or weak).
Recommendation
Main recommendation
In elderly (75 years or older) patients living in Latin America with severe symptomatic aortic stenosis candidates for transfemoral approach, the panel suggests the use of TAVI over SAVR. This is a conditional recommendation, based on moderate certainty of the evidence (⨁⨁⨁Ο). A summary of the main recommendation is presented in figure 2 and the visual summary in online supplemental file 1.
Supplemental material
Main recommendation. The suggested age threshold of 75 years old should be used as guidance rather than a determinant threshold. Patients discussed in Heart team with factors favouring SAVR may result in a surgical preferred course of action, while the presence of favourable aspects for TAVI would further support the recommendation for TAVI. SAVR, surgical aortic valve replacement; TAVI, transcatheter aortic valve implant.
Remarks
The suggested age threshold of 75 years old is based on the judgement of limited available literature and should be used as guidance rather than a determinant threshold. When implementing the recommendation, several clinical aspects need to be assessed by a comprehensive clinical group, ideally through a Heart Team discussion (table 1). Patients with factors favouring SAVR may result in a surgical preferred course of action, especially in the lower risk and younger age ranges. On the other hand, the presence of favourable aspects for TAVI would further support the recommendation for TAVI. Although these clinical circumstances favouring TAVI mostly apply to 75-year-old patients, they may also apply to younger candidates under specific circumstances.
Subgroup considerations
Summary of the evidence
We identified eight randomised trials including 7961 patients, 4087 treated with TAVI and 3874 patients treated with SAVR (tables 2 and 3). The vast majority of the patients included in these trials were elderly people with transfemoral TAVI. Two trials were not sponsored by the industry for regulatory approval.10 13
Summary of findings at 30 days
Summary of findings at longest-term
Benefits and harms
The overall judgement was that the benefits and harms of TAVI were considered moderate and small, respectively (figure 3, details presented in the online supplemental material). At 30 days, there were no statistically significant differences for perioperative mortality (relative risk (RR) 0.78, 95% CI 0.58 to 1.05; 6 fewer per 1000, from 11 fewer to 1 more), cardiovascular mortality (RR 0.81, 95% CI 0.59 to 1.10; 4 fewer per 1000, from 9 fewer to 2 more), or stroke (RR 0.75, 95% CI 0.57 to 1.00; 11 fewer per 1000, from 20 fewer to 0 fewer). TAVI was associated with lower risk of life-threatening bleeding (RR 0.27, 95% CI 0.15 to 0.46; 176 fewer per 1000, from 205 fewer to 130 fewer), rehospitalisations (RR 0.75, 95% CI 0.58 to 0.97; 11 fewer per 1000, from 19 fewer to one fewer), atrial fibrillation (RR 0.26, 95% CI 0.20 to 0.34; 256 fewer per 1000, from 276 fewer to 228 fewer) and better quality of life (RR 1.52 for a 5-point improve, 95% CI 1.34 to 1.74; 258 more per 1000, from 169 more to 368 more) but higher pacemaker risk (RR 2.32, 95% CI 1.51 to 3.56; 77 more per 1000, from 30 more to 150 more), vascular complications (RR 2.75, 95% CI 1.79 to 4.22; 40 more per 1000, from 18 more to 73 more) and paravalvular leak (RR 5.70, 95% CI 2.95 to 10.99; 33 more per 1000, from 14 more to 69 more).
Summary of judgements from the evidence-to-decision framework. Each row represents a different criterion or domain from the evidence-to-decision discussion with the panel. Conclusions of each domain arise from consensus or voting. Blue and green represent judgements favouring TAVI, while orange or red represents judgements not favouring TAVI. SAVR, surgical aortic valve replacement; TAVI, transcatheter aortic valve implant.
Based on the longest follow-up data available from each trial, there were no statistically significant differences for mortality (RR 0.97, 95% CI 0.90 to 1.04; 8 fewer per 1000, from 25 fewer to 10 more), cardiovascular mortality (RR 0.94, 95% CI 0.84 to 1.06; 14 fewer per 1000, from 38 fewer to 14 more) or stroke (RR 0.94, 95% CI 0.81 to 1.10; 9 fewer per 1000, from 28 fewer to 15 more). TAVI was associated with more reinterventions (RR 2.34, 95% CI 1.15 to 4.75; 7 more per 1000, from 1 more to 19 more) but similar rehospitalisations (RR 1.00, 95% CI 0.78 to 1.29; 0 fewer per 1000, from 67 fewer to 88 more) and quality of life (RR 1.03 for a 5-point improve, 95% CI 1.00 to 1.06; 23 more per 1000, from 0 fewer to 46 more).
Certainty of the evidence
The certainty of the evidence across outcomes was considered moderate. Critical outcomes were judged as high certainty, while the majority of important outcomes as moderate. Thirty-day patient–prosthesis mismatch, long-term rehospitalisations and reinterventions were rated as low certainty, none of them considered critical for decision-making.
Although there was imprecision to support a reduction of critical outcomes (mortality, cardiovascular mortality and stroke), the methodological team considered that there was high certainty to rule out a risk increase beyond the prespecified minimal clinically important difference. Other common reasons to downgrade the certainty of the evidence in important but not critical outcomes were open-label design, ascertainment bias and heterogeneous outcome definitions.
Values and preferences, acceptability, feasibility and resource use
The panel considered that there was no important heterogeneity or uncertainty in patients’ preferences between the two interventions based on low to very low quality of evidence. They also stated that TAVI would be acceptable by patients and stakeholders and feasible to implement in Latin America. The major concern from panellists was related to upfront TAVI costs in resource-constrained settings.
Balance of effects
The whole panel voted in favour of TAVI in this elderly population. Key reasons for the direction of this recommendation were proven superiority or the non-inferiority of many patient-important outcomes while being a less invasive procedure, implementable and accepted intervention in our region. The main reason for a conditional instead of a strong recommendation was the concern of increased upfront costs for a resource-constrained region. The panel also highlighted that the benefit of TAVI over SAVR was not large for some patient-important outcomes, especially in low-risk patients using the presented age threshold.
Considerations for implementation
Dealing with conditional recommendations highlights the importance of shared decision-making process with patients.24 Conditional recommendations imply that different choices will be appropriate for different patients putting a higher value on patient values and preferences. Physicians must help each patient to arrive at a consistent treatment decision.
The role of local resource availability to cover TAVI costs plays a major role in this region. Stakeholders with responsibility for cost distribution (healthcare insurances, government and hospital administrators, among others) should make an effort to consolidate the decision of this panel with the reality of resource use of their specific settings. Potential strategies to increase the affordability of TAVI in Latin America include reduced or waived taxation, or large bill contracts for health insurances.
TAVI programmes are at different stages of development in Latin America. Each centre should have a tailored plan to guarantee that clinical results of both TAVI and SAVR are optimised to resemble as much as possible the results of supporting evidence from high-income countries. Measurement, monitoring and consideration of local expertise and results should be taken into account when applying this recommendation in Latin America. More details about the implementation, future research priorities, update process, dissemination and monitoring are detailed in the online supplemental material.
Ethics statements
Acknowledgments
We thank all the participants of this guideline process, including both postsurgical aortic valve replacement and post-transcatheter aortic valve implant (TAVI) patients (names not disclosed in this report), Monica Buus and Laura Stura as assistants, Dr William Toff from for spending time clarifying valuable information about UK TAVI trial execution and results, and external reviewers Marcelo Trivi, Oscar Mendiz and Carlos Rivas for improving the quality of the document.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Twitter @lamelaspablo, @Hbribeiro37, @fernando_wyss
Correction notice In response to reader queries, the authors have made corrections to their article since it was first published online. The authors have now expanded on the role of SOLACI and the involvement of other societies in this guideline. Sentences have been added to the 'Conflict of interest management and role of the sponsor' section, and a new section has been added entitled 'Involvement of other societies'. Conflicts of interest for external clinical reviewers have now been detailed in the 'Competing interests' section.
Contributors PL, MAR, IJ and AI developed and oversight the whole process of the guideline. RB and HR collaborated as topic experts. AB, FW, ASL, GO, MA, JCR, JC, CA, FC, AM and FSB collaborated as panelists. JAM, AD, OS and PH approved from Sociedad Latino Americana de Cardiología Intervencionista in the quality assurance process. RN collaborated in the design phase and then made major contributions during the review.
Funding Sociedad Latino Americana de Cardiología Intervencionista (SOLACI) covered costs related to the systematic review and data abstraction. SOLACI authorities had no involvement in the process of the guideline until quality assurance following the external review.
Competing interests FSB: proctoring fees from Edwards LifeSciences, Medtronic and Boston Scientific. CA, MA, FC: proctoring fees from Edwards LifeSciences and Boston Scientific. Participants without voting rights: PL: proctoring fees from Edwards LifeSciences, Medtronic, Boston Scientific (significant); consulting fees Boston Scientific (not significant). HR: proctoring fees from Edwards LifeSciences, Boston Scientific and Medtronic, and research grants from Edwards LifeSciences, Boston Scientific and Medtronic. Clinical external reviewers: MT: Nothing to declare. OM: Consulting honoraria from Medtronic, Boston Scientific and Philips. CR: Nothing to declare.
Patient and public involvement statement Patients were included in the guideline panel
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.