The developers1 of the first rigorous predictive model for mortality
after transcatheter aortic valve implantation (TAVI) have overcome the
limitations of the previous surgical scores. While EuroSCORE I is an old
and redundant model based on data of 1995 and derived from a highly
heterogeneous patient group with different operations, techniques and
demographics, this predictive model is based on new results of a
remarkably homogeneous population (72,4% >80 years of age and all
treated with TAVI procedures).1 Moreover, while EuroSCORE II developers
included variables that were not significantly associated with the event
by multivariate regression and did not analyze some important variables2,
Iung et al1 studied a wide variety of variables including only those
significantly associated with mortality. In addition, while in traditional
cardiac surgery there is a great disparity in results between centers
which can hamper the accuracy of any predictive model, in this study there
was no difference according to the center.1
However, and despite all these strengths, the discriminatory power
evaluated by C-index in the validation cohort was the same as that
observed when the old EuroSCORE was used (0,59). This disappointing result
means that the ability to distinguish or separate those who will die after
TAVI from those who will not is the same between a score created 15 years
ago to predict mortality after traditional surgery and a score created
nowadays and based on patients who underwent TAVI. In fact, taking into
account that a C-statistic of 0.5 indicates no predictive ability, a value
of 0,59 does not seem to be very favorable. This is more pronounced when
we consider that the promising results published by the authors of the
EuroSCORE II (C-statistic of 0,81) proved again to be disappointing when
the external validation of the model was assessed by independent
researchers2.
We agree with Ribeiro and Rod?s-Cabau3 when noting that this new model is
more accurate than some surgical scores. However, EuroSCORE II has shown
in some studies4 great calibration and better discrimination power (C-
statistic of 0,66). These authors consider the smaller number of patients
compared with surgical scores as a reason that could explain this enormous
limitation. We complete agree with this suggestion and we would like to
note that, while this system has been created based on a multivariate
analysis of 28 variables (Table 2, those with p<0,2) in a development
group of 2552 patients (253 deaths), EuroSCOCRE II was created using 26
variables in 12673 patients (587 deaths). Moreover, these authors wonder
whether a predictive model for octogenarian patients would be more
reliable taking into account a medium-term follow up. This suggestion is
supported by some works that have already confirmed this finding in
traditional surgery.5
Therefore, many more patients should be included in future models and the
performance of the system at mid-term follow up should be always tested.
However, the creation of an efficient and reliable predictive model for
TAVI seems to be one of the biggest challenges facing cardiologist and
cardiac surgeons.
REFERENCES
1. Iung B, Laou?nan C, Himbert D et al. Predictive factors of early
mortality after transcatheter aortic valve implantation: individual risk
assessment using a simple score. Heart. 2014 Apr 16. doi: 10.1136/heartjnl
-2013-305314.
2. Barili F, Pacini D, Capo A et al. Does EuroSCORE II perform better than
its original versions? A multicentre validation study. Eur Heart J
2013;34:22-9.
3. Ribeiro HB, Rod?s-Cabau J. The multiparametric FRANCE-2 risk score: one
step further in improving the clinical decision-making process in
transcatheter aortic valve implantation. Heart. 2014 Apr 23. doi:
10.1136/heartjnl-2014-305806.
4. Durand E, Borz B, Godin M et al. Performance Analysis of EuroSCORE II
Compared to the Original Logistic EuroSCORE and STS Scores for Predicting
30-Day Mortality After Transcatheter Aortic Valve Replacement. Am J
Cardiol 2013;111:891-7.
5. Leontyev S, Walther T, Borger MA et al. Aortic valve replacement in
octogenarians: utility of risk stratification with EuroSCORE. Ann Thorac
Surg. 2009;87:1440-5.
Conflict of Interest:
None declared
The developers1 of the first rigorous predictive model for mortality after transcatheter aortic valve implantation (TAVI) have overcome the limitations of the previous surgical scores. While EuroSCORE I is an old and redundant model based on data of 1995 and derived from a highly heterogeneous patient group with different operations, techniques and demographics, this predictive model is based on new results of a remarkably homogeneous population (72,4% >80 years of age and all treated with TAVI procedures).1 Moreover, while EuroSCORE II developers included variables that were not significantly associated with the event by multivariate regression and did not analyze some important variables2, Iung et al1 studied a wide variety of variables including only those significantly associated with mortality. In addition, while in traditional cardiac surgery there is a great disparity in results between centers which can hamper the accuracy of any predictive model, in this study there was no difference according to the center.1 However, and despite all these strengths, the discriminatory power evaluated by C-index in the validation cohort was the same as that observed when the old EuroSCORE was used (0,59). This disappointing result means that the ability to distinguish or separate those who will die after TAVI from those who will not is the same between a score created 15 years ago to predict mortality after traditional surgery and a score created nowadays and based on patients who underwent TAVI. In fact, taking into account that a C-statistic of 0.5 indicates no predictive ability, a value of 0,59 does not seem to be very favorable. This is more pronounced when we consider that the promising results published by the authors of the EuroSCORE II (C-statistic of 0,81) proved again to be disappointing when the external validation of the model was assessed by independent researchers2. We agree with Ribeiro and Rod?s-Cabau3 when noting that this new model is more accurate than some surgical scores. However, EuroSCORE II has shown in some studies4 great calibration and better discrimination power (C- statistic of 0,66). These authors consider the smaller number of patients compared with surgical scores as a reason that could explain this enormous limitation. We complete agree with this suggestion and we would like to note that, while this system has been created based on a multivariate analysis of 28 variables (Table 2, those with p<0,2) in a development group of 2552 patients (253 deaths), EuroSCOCRE II was created using 26 variables in 12673 patients (587 deaths). Moreover, these authors wonder whether a predictive model for octogenarian patients would be more reliable taking into account a medium-term follow up. This suggestion is supported by some works that have already confirmed this finding in traditional surgery.5 Therefore, many more patients should be included in future models and the performance of the system at mid-term follow up should be always tested. However, the creation of an efficient and reliable predictive model for TAVI seems to be one of the biggest challenges facing cardiologist and cardiac surgeons.
REFERENCES 1. Iung B, Laou?nan C, Himbert D et al. Predictive factors of early mortality after transcatheter aortic valve implantation: individual risk assessment using a simple score. Heart. 2014 Apr 16. doi: 10.1136/heartjnl -2013-305314. 2. Barili F, Pacini D, Capo A et al. Does EuroSCORE II perform better than its original versions? A multicentre validation study. Eur Heart J 2013;34:22-9. 3. Ribeiro HB, Rod?s-Cabau J. The multiparametric FRANCE-2 risk score: one step further in improving the clinical decision-making process in transcatheter aortic valve implantation. Heart. 2014 Apr 23. doi: 10.1136/heartjnl-2014-305806. 4. Durand E, Borz B, Godin M et al. Performance Analysis of EuroSCORE II Compared to the Original Logistic EuroSCORE and STS Scores for Predicting 30-Day Mortality After Transcatheter Aortic Valve Replacement. Am J Cardiol 2013;111:891-7. 5. Leontyev S, Walther T, Borger MA et al. Aortic valve replacement in octogenarians: utility of risk stratification with EuroSCORE. Ann Thorac Surg. 2009;87:1440-5.
Conflict of Interest:
None declared