Issued in 2016 – Percutaneous Valve Registry

Percutaneous Valve Registry Annual report 2015

Issued in 2016 – Percutaneous Valve Registry Working group PercutaneousValve Registry 2015 Chair Andreas Rück, MD, PhD, Stockholm Deputy chair Henrik Bjursten, MD, PhD, Lund Stefan James, MD, PhD, Uppsala Kenneth Mier, MD, PhD, Karlskrona Niels Erik Nielsen, MD, Linköping Johan Nilsson, MD, PhD, Umeå Petur Petursson, MD, Gothenburg Ninos Samano, MD, Örebro Magnus Settergren, MD, PhD, Stockholm

Issued in 2016 – Percutaneous Valve Registry Figure 1. TAVI Quality Index, per center, The TAVI Quality Index, now published for the second time, is intended to encourage quality improvements by benchmarking. Maximum score is 10 points. Three structural factors are scored, with a maximum of 5 points in total. Multidisciplinary heart team and cardiac surgery on-site was present at all hospitals. Procedural volume varied, this year two hospitals performed less than 50 procedures, and one hospital performed more than 100 procedures. For the results, another maximum 5 points are assigned for the following factors: major complications (major vascular complications, major bleeding, stroke) and risk-adjusted mortality (30-day mortality adjusted for EuroSCORE II). Results have improved in many hospitals compared with last year. Detailed results are found in Table 1. Overall, the TAVI Quality Index ranged from 6 to 8 points, with Linköping and Örebro having the highest score. Notably, three centers (Linköping, Örebro and Karlskrona) did not provide data on complications to allow for adjudication, thus the bars showing the complication data are striped in Figure 1. For remaining centers, we have centrally adjudicated all cases for major complications (major vascular complications, major bleeding, and stroke) against source documents, which will ensure the highest possible data quality for these fields in the Quality Index. Detailed definitions of the different parts of the TAVI QI are provided in the online appendix at

Issued in 2016 – Percutaneous Valve Registry Figure 2. Number of TAVI procedures, per centre and year, 2008–2015. The total number of TAVI procedures increased by 34 % to 577 in Details on the numbers of procedures can also be found in Table 1. The largest number of cases was at Karolinska (155 cases) while the lowest number of cases was at the recently started TAVI service in Karlskrona (12 cases).

Issued in 2016 – Percutaneous Valve Registry Figure 3. Number of TAVI procedures, per inhabitants, The number of TAVI procedures per inhabitants in the different counties of Sweden still shows marked differences. For the less populated counties, some of the differences might be random fluctuations due to low case numbers. Stockholm (Karolinska), Gothenburg (Sahlgrenska) and Skåne (Lund) regions have over 1 million inhabitants each. Among these three, the Gothenburg region has a much lower TAVI rate than the other two. This is probably due to economic constraints and not differences in the prevalence of aortic stenosis. Data is calculated on patients’ place of residency, not on the hospital completing the procedure.

Issued in 2016 – Percutaneous Valve Registry Figure 4. Mean age of patients, per centre and year, 2009–2015. The mean age of the patients is still over 80 years, with no major change over the last years. This indicates that there has been no major change in the selection of patients between TAVI and surgical aortic valve replacement. Probably a larger proportion of the elderly patients with aortic stenosis have not been offered any aortic valve replacement previously.

Issued in 2016 – Percutaneous Valve Registry Figure 5. Logistic EuroSCORE I, per centre and year, 2009–2015. Logistic EuroSCORE is still high and mean score has not changed much in the last years. This implies that among aortic stenosis patients, TAVI is offered mostly to high-risk patients, whereas the lower risk patients are still treated surgically. The logistic EuroSCORE is a risk scoring system developed to estimate perioperative mortality associated with cardiac surgery. It is used here to estimate the severity of comorbidities, although there is evidence that the logistic EuroSCORE generally overestimates mortality risks. Logistic EuroSCORE exists in two versions: The original logistic EuroSCORE I, which has been included in this registry from the beginning in 2008, and the revised EuroSCORE II, which was added in In this report, the EuroSCORE II is used for risk- adjustment in the TAVI QI (Table 1). Logistic EuroSCORE I is used here for historical comparisons.

Issued in 2016 – Percutaneous Valve Registry Figure 6. Mean fluoroscopy time, per type of device and year, 2008–2015. The mean fluoroscopy time has plateaued at about 20 minutes. The variation between different valve types seems small. Fluoroscopy time can be seen as a marker for technical complexity of the procedure.

Issued in 2016 – Percutaneous Valve Registry Figure 7. Mean contrast agent volume per procedure (mL), per centre and year, 2008–2015. The volume of contrast agent used per procedure decreased somewhat from last year. There was a large variation between hospitals, with Linköping using a mean volume of only 9 mL of contrast. High volume of contrast agent use can deteriorate kidney function in susceptible patients. In that regard it is reassuring that 6 out of 8 hospitals had a mean volume contrast use of less than 100 mL per procedure, which generally is a safe amount.

Issued in 2016 – Percutaneous Valve Registry Figure 8. Number of TAVI procedures, per access type and year, 2008–2015. The transfemoral access route is the predominant choice to a similar extent as last year (87 %). Alternative (non-femoral) access routes include transapical, subclavian, and transaortic access. There is a trend to smaller profile transfemoral TAVI systems, and therefore the need for non-femoral access is predicted to decrease for TAVI. Non-femoral access usually requires general anaesthesia, while most transfemoral cases are performed under local anaesthesia only.

Issued in 2016 – Percutaneous Valve Registry Figure 9. Proportion and number of TAVI procedures, per access type and centre, Despite treating similar patients, the proportion of alternative (non-femoral) access cases is largely different between hospitals. Sahlgrenska and Karlskrona used 100 % femoral access, while Umeå used femoral access in 68 %. A possible explanation is that the threshold for femoral versus non-femoral access is at a different femoral artery diameter in the centers, depending on which transfemoral TAVI system is used. Presently the Medtronic EvolutR has the lowest profile on the market.

Issued in 2016 – Percutaneous Valve Registry Figure 10. Proportion and number of TAVI procedures, per type of device and centre, The main TAVI devices used were the Medtronic CoreValve/EvolutR and the Edwards Sapien. The Boston Lotus was the third most frequent device, while the Symetis Acurate neo was used in a smaller number of patients in one hospital. Different generations of the Medtronic devices (Corevalve and EvolutR) and the Edwards Sapien (XT and S3) are grouped together in this report.

Issued in 2016 – Percutaneous Valve Registry Figure 11. Proportion of aortic regurgitation severity assessed by postoperative echocardiography, 2008–2015. Number of patients above the bar. Aortic regurgitation is seen in a proportion of patients after TAVI. Usually this regurgitation is paravalvular, caused by incomplete sealing between the patient’s calcified aortic annulus and the prosthesis. There are no clear differences between the valve types. Reported data is acquired by transthoracic echocardiography one to five days after the procedure. Data is site-reported, without any core-lab assessment of the images. As the grading of paravalvular regurgitation by echocardiography often is difficult, it would be problematic to draw firm conclusions from minor differences in site-reported data for this parameter.

Issued in 2016 – Percutaneous Valve Registry Figure 12. Proportion and number and proportion of patients with need for a new permanent pacemaker after TAVI, per device and year, 2008–2015. Aortic valve implantation causes atrioventricular block and bradycardia in a proportion of patients. This is due to pressure from the valve prosthesis on the patient’s conduction system. The Medtronic Corevalve/EvolutR still has the highest pacemaker rate, now at 16 %. The Edwards Sapien has a low pacemaker rate, as has also been noted in other studies. Although the numbers are limited, the Symetis Acurate seems to have a very low pacemaker rate (0 % in 22 patients). The Boston Lotus had a pacemaker rate of 12 %, higher than some other valves in our registry, but lower than most published data.

Issued in 2016 – Percutaneous Valve Registry Figure 13. Proportion and number of patients with in-hospital stroke, per year, 2008–2015. Frequency of in-hospital stroke is still about 2 % without any major changes in the last years.

Issued in 2016 – Percutaneous Valve Registry Figure 14. Thirty-day mortality after TAVI, per year, 2008–2015. Number above the bar. Thirty-day mortality has decreased substantially in 2015, now only 2.3 %. It is not clear what has caused this decrease from over 6 % in It might be that the highest risk cases were considered not eligible for TAVI, although the average risk profile and age has not declined. Mortality data is from the national population statistics, which has complete coverage.

Issued in 2016 – Percutaneous Valve Registry Figure 15. One-year mortality after TAVI, per year, 2008–2014. Number above the bar. The one-year mortality was 16 % for patients treated in 2014, an increase from the previous year. One-year mortality is largely dependent on the patient selection. Probably the patients treated in 2015 will have a lower one- year mortality, since their 30-day mortality was very low (Figure 14).

Issued in 2016 – Percutaneous Valve Registry Figure 16. Five-year survival after TAVI according to Kaplan-Meier, per age group, 2008–2015. Long-term survival after TAVI is only slightly worse with higher age. The probable reason is that the patient selection is stricter with higher age. Total five-year survival is about 50 %. Approximately 400 patients had their TAVI procedure done at least five years ago.

Issued in 2016 – Percutaneous Valve Registry Figure 17. Five-year survival after TAVI according to Kaplan-Meier, per kidney function group, 2008–2015. Long-term survival after TAVI is considerably worse in patients with decreased kidney function. Still, the five- year survival was about 30 % in the group with the poorest kidney function. Glomerular filtration rate (GFR) was calculated according to Cockcroft-Gault.

Issued in 2016 – Percutaneous Valve Registry Table 1. TAVI Quality Index, per centre, Detailed data for the TAVI Quality Index for all hospitals. See Figure 1 for further discussion. Thirty-day mortality and EuroSCORE II data are the last two years combined, to decrease the effect of random fluctuations in centers with low annual case numbers.Linköping, Karlskrona and Örebro did not provide data for adjudication of complications, thus their complication data is only self-reported.

Issued in 2016 – Percutaneous Valve Registry Table 2. Number of TAVI procedures per centre and year, 2008–2015. Exact case numbers for all centers for all years. See also figure 2.

Issued in 2016 – Percutaneous Valve Registry Table 3. Number of TAVI procedures, per access type and year, 2008–2015. Alternative (non-femoral) access is used in only a minority of cases. See also Figure 8.

Issued in 2016 – Percutaneous Valve Registry Table 4. Number (proportion in brackets) of selected intra-procedurale complications, per type of device, Life-threatening intra-procedural complications occur infrequently, but still emergency surgery and cardiopulmonary bypass are necessary in about 2 % of cases. The Medtronic CoreValve system led to the use of more than one prosthesis in a high number of cases. This is expected to decrease substantially in 2016 when this product has been replaced by the Medtronic EvolutR. The number of vascular complications with the Boston Lotus tended to be higher than with other products, maybe related to the large diameter femoral access needed. Most other complications do not seem to be associated with a particular valve type.

Issued in 2016 – Percutaneous Valve Registry Table 5. Number (proportion in brackets) of selected in-hospital (post- procedural) complications, per type of device, The number of vascular and bleeding complications is decreasing. Need for new permanent pacemaker is higher for certain valves (Medtronic CoreValve, Boston Lotus). The number of strokes is rather low, but this complication still occurs in 2 %. The most important severe complications (major vascular, major bleeding, and stroke) were centrally ­ adjudicated against source data to exclude any inconsistency. Some centers (Linköping, Örebro, Karlskrona) did however not provide source data for adjudication.

Issued in 2016 – Percutaneous Valve Registry Table 6. Summary of 30-day and one-year mortality and EuroSCORE I, per center, 2015.