Open Issues in TAVI Part 1
Open Issues in TAVI Part 1
The decision of surgical aortic valve replacement (SAVR) or TAVI in symptomatic severe AS depends on the presence of contraindications for SAVR or a high-surgical risk. Two risk scores are used to calculate the risk of cardiac surgery (including SAVR): the Society of Thoracic Surgeons-Predicted Risk of Mortality score (STS-PROM) and the European System for Cardiac Operative Risk Evaluation (EuroSCORE) model (additive and logistic). When the STS-PROM score exceeds 10% or when the logistic EuroSCORE is ≥20%, referral for TAVI should be considered. These risk scores, however, were developed and validated in the standard surgical risk populations and predict short-term mortality (hospital or 30-day) after cardiac surgery. Both scoring systems have, therefore, important deficiencies especially in a heterogeneous high-risk patient population such as TAVI candidates. Moreover, they are inadequate to predict the long-term outcome or procedural complications. The EuroSCORE II has recently been developed from data of 22 381 patients undergoing cardiac surgery and validated in a subset of 5553 patients. Variables such as impaired mobility, New York Heart Association (NYHA) functional class and diabetes were incorporated into the new risk score. This score appears to have superior discriminatory power for predicting 30-day mortality over the original logistic EuroSCORE.
The reasons for suboptimal performance of the scores are diverse. The development cohort of patients from whom a score was derived is usually very different from the patients to whom it is applied. Also, changes in surgical techniques, peri-operative care, patients' characteristics, and differences in hospitals and even surgeons or interventionalists may further explain the modest performance of the scores. Finally, risk scores are usually developed through the standard statistical approaches, not taking into account risk factor interactions or procedure-specific weightings.
The application of risk scores for referral to TAVI is also limited, since conventional cardiovascular risk factors (i.e. peripheral vascular disease, diabetes) are included in the scores, but specific risk factors for TAVI are not included, such as frailty, porcelain aorta, vessel tortuosity, chest wall malformation, or chest radiation. In addition, risk predictions should be based on standardized definitions such as the Valve Academic Research Consortium (VARC) criteria.
The EuroSCORE II showed better discriminatory power for predicting 30-day mortality after TAVI, when compared with the logistic EuroSCORE and STS-PROM in a cohort of 350 TAVI patients (areas under the curve 0.70, 0.61, and 0.59, respectively). Recently, based on data from the FRANCE-2 registry, a risk score has been proposed which includes a specific procedural variable in the calculation: the TAVI access site (transfemoral vs. non-transfemoral). However, the discrimination of this risk score was still modest (area under the curve 0.59), indicating the limitations of the score to reliably individualize risk.
Accordingly, the unmet needs currently include improvements in risk scores to better individualize patient risk in the populations considered for TAVI. The open questions are (i) how to improve discrimination between low- and high-surgical risk (including SAVR), (ii) how to identify patients who should be referred for TAVI, and (iii) how to predict procedural risk and outcome in TAVI candidates.
It may be difficult (if not impossible) to develop the specific scores that answer these questions, since this concerns a complex patient group, in whom individualized decision-making may be preferred to ensure that every patient receives the optimal therapy. In addition to scores, discussion on individual patient management should take place in the 'Heart Team' including cardiologists with specific interest on valvular heart disease, transcatheter interventions and cardiac imaging, cardiothoracic surgeons, anaesthesiologists, and other specialists such as geriatricians.
Risk Scores for Patient
The decision of surgical aortic valve replacement (SAVR) or TAVI in symptomatic severe AS depends on the presence of contraindications for SAVR or a high-surgical risk. Two risk scores are used to calculate the risk of cardiac surgery (including SAVR): the Society of Thoracic Surgeons-Predicted Risk of Mortality score (STS-PROM) and the European System for Cardiac Operative Risk Evaluation (EuroSCORE) model (additive and logistic). When the STS-PROM score exceeds 10% or when the logistic EuroSCORE is ≥20%, referral for TAVI should be considered. These risk scores, however, were developed and validated in the standard surgical risk populations and predict short-term mortality (hospital or 30-day) after cardiac surgery. Both scoring systems have, therefore, important deficiencies especially in a heterogeneous high-risk patient population such as TAVI candidates. Moreover, they are inadequate to predict the long-term outcome or procedural complications. The EuroSCORE II has recently been developed from data of 22 381 patients undergoing cardiac surgery and validated in a subset of 5553 patients. Variables such as impaired mobility, New York Heart Association (NYHA) functional class and diabetes were incorporated into the new risk score. This score appears to have superior discriminatory power for predicting 30-day mortality over the original logistic EuroSCORE.
The reasons for suboptimal performance of the scores are diverse. The development cohort of patients from whom a score was derived is usually very different from the patients to whom it is applied. Also, changes in surgical techniques, peri-operative care, patients' characteristics, and differences in hospitals and even surgeons or interventionalists may further explain the modest performance of the scores. Finally, risk scores are usually developed through the standard statistical approaches, not taking into account risk factor interactions or procedure-specific weightings.
The application of risk scores for referral to TAVI is also limited, since conventional cardiovascular risk factors (i.e. peripheral vascular disease, diabetes) are included in the scores, but specific risk factors for TAVI are not included, such as frailty, porcelain aorta, vessel tortuosity, chest wall malformation, or chest radiation. In addition, risk predictions should be based on standardized definitions such as the Valve Academic Research Consortium (VARC) criteria.
The EuroSCORE II showed better discriminatory power for predicting 30-day mortality after TAVI, when compared with the logistic EuroSCORE and STS-PROM in a cohort of 350 TAVI patients (areas under the curve 0.70, 0.61, and 0.59, respectively). Recently, based on data from the FRANCE-2 registry, a risk score has been proposed which includes a specific procedural variable in the calculation: the TAVI access site (transfemoral vs. non-transfemoral). However, the discrimination of this risk score was still modest (area under the curve 0.59), indicating the limitations of the score to reliably individualize risk.
Accordingly, the unmet needs currently include improvements in risk scores to better individualize patient risk in the populations considered for TAVI. The open questions are (i) how to improve discrimination between low- and high-surgical risk (including SAVR), (ii) how to identify patients who should be referred for TAVI, and (iii) how to predict procedural risk and outcome in TAVI candidates.
It may be difficult (if not impossible) to develop the specific scores that answer these questions, since this concerns a complex patient group, in whom individualized decision-making may be preferred to ensure that every patient receives the optimal therapy. In addition to scores, discussion on individual patient management should take place in the 'Heart Team' including cardiologists with specific interest on valvular heart disease, transcatheter interventions and cardiac imaging, cardiothoracic surgeons, anaesthesiologists, and other specialists such as geriatricians.
