New York Heart Association (NYHA) class IV heart failure is one of the factors used in predicting in-hospital mortality in patients undergoing transcatheter aortic valve replacement (TAVR). The effect of systolic heart failure (SHF), aside from NYHA classification, on peri-procedural outcomes is unclear.
The study population was identified from the 2016 Nationwide Readmissions Data database using International Classification of Diseases-Tenth Revision codes for TAVR and SHF. Study end points included in-hospital all-cause mortality, the length of hospital stay, cardiogenic shock, acute myocardial infarction (AMI), acute kidney injury (AKI), mechanical complications of prosthetic valve, bleeding, and 30-day readmission rate. Propensity matching was used to create a control group of TAVR patients without a SHF diagnosis (TAVR-C).
A total of 5,674 patients were included in each group (mean age 79.9 years; 35.6% female). The groups were comparable in terms of baseline characteristics and comorbidities. TAVR-SHF was associated with significantly higher in-hospital all-cause mortality (2.7% vs. 1.9%, p<0.01), longer hospital stay (7.5 vs. 5.5 days, p<0.01), higher cardiogenic shock (5.1% vs. 1.6%, p<0.01), AMI (4.0% vs. 1.9%, p<0.01), AKI (18.7% vs. 12.4%, p<0.01) and mechanical complications of prosthetic valve (1.2% vs. 0.6%, p<0.01). There was no significant difference between TAVR-SHF and TAVR-C in terms of bleeding (19.5% vs. 18.2%, p=0.08) and 30-day readmission rate (10.8% vs. 10.2%, p=0.29).
Compared with TAVR-C, TAVR-SHF was associated with higher in-hospital peri-procedural complications and all-cause mortality.
Introduction
Systolic heart failure (SHF) in patients with severe aortic stenosis (AS) carries a worse prognosis, and aortic valve replacement improves ventricular systolic function and survival.1,2 Therefore, SHF is an indication for aortic valve replacement in severe AS.2 Both surgical aortic valve replacement (SAVR) and transcatheter aortic valve replacement (TAVR) are associated with comparable survival and ventricular systolic function recovery in this group of patients.3 TAVR, however, is the recommended approach for patients with intermediate to prohibitive surgical risk; and SHF patients are often considered a high-risk group. As a result, SHF represents a significant proportion of the TAVR population.4,5
The Society of Thoracic Surgeons (STS) and the European System for Cardiac Operative Risk Evaluation (EuroSCORE II) are the most commonly used risk score calculators to predict peri-operative mortality in cardiac surgery.6 Both scores consider ventricular function during peri-operative risk assessment for SAVR, which is highly dependent on pre-operative systolic function; however, studies have shown conflicting data on post-TAVR outcomes with pre-existing SHF, and none of these scores provide an acceptable predictive ability for post-TAVR outcomes.5,7,8
Method
Data source
The Nationwide Readmissions Data (NRD) is one of the largest collections of de-identified longitudinal hospital care data in the US. The NRD supports various types of analyses, including readmission rates, with safeguards to protect the privacy of individual patients, physicians, and hospitals. It contains clinical and nonclinical variables for each hospital stay, including patient linkage number for linking hospital visits across hospitals, International Classification of Diseases, Tenth Revision, Clinical Modification/Procedure Coding System (ICD-10-CM/PCS) for principal and secondary procedures and diagnoses, age, gender, length of stay (LOS), and others.9,10
Study cohort
The ICD-10-CM/PCS codes were used to search discharges in the 2016 NRD who had TAVR during the index hospitalisation; SHF, baseline characteristics, comorbidities, and in-hospital post-procedural complications were subsequently extracted.10 The NRD excludes discharges with missing age, missing linkage numbers or from hospitals with more than 50% of their discharges excluded because of these criteria, as patients treated in these hospitals may not be reliably tracked over time.9 All best practices to use the NRD highlighted by Khera et al. were followed.11
Study end points
The study end points included in-hospital all-cause mortality, length of index hospital stay (LOS), cardiogenic shock, acute myocardial infarction (AMI), acute kidney injury (AKI), mechanical complications of prosthetic valve, bleeding, and 30-day readmission rate. Mechanical complications of prosthetic valve defined as TAVR valve embolisation, displacement, breakdown, or other mechanical complications (excluding peri-valvular leak, fibrosis, infection, stenosis, and thrombosis). AMI included any new post-procedural subendocardial or transmural myocardial infarction. AKI included any new post-procedural acute worsening of kidney function. Bleeding included any circulatory or central nervous system bleeding during or post-procedure, or post-haemorrhage/procedure anaemia.
The 30-day readmission rate was calculated based on the NRD recommendations. We identified all-cause readmissions (including first and subsequent admissions) within 30 days post-discharge to any hospital within the same state (as cross-state readmissions cannot be tracked by the NRD database). Transfers were not considered readmissions. We excluded TAVR patients whose age was less than 18 years, missing LOS or if TAVR procedure was done in the month of December as readmission rate cannot be calculated in two different years.12
Statistical analysis
Statistical Analysis System (SAS) software 9.4 (TS1M4, SAS Institute Inc., Cary, North Carolina) was used for propensity score matching and statistical analysis. The logistic regression was used to create propensity score, based on baseline characteristics for one-to-one parallel, balanced propensity score matching model using a caliper of 0.05. McNemar test was used to compare paired categorical variables, while paired-samples t-test was used to compare continuous variables. Pearson Chi-Square of Independence and unpaired-sample t-test were used in subgroup analysis to compare the end points of interest in TAVR-SHF to the overall TAVR group before propensity matching. The multi-variable logistic regression model was used to identify predictors of in-hospital all-cause mortality of the TAVR-SHF group. A two-tailed p value of <0.05 was used for statistical significance.13,14
Results
In the 2016 NRD database, there were around 17.2 million discharges. There were 23,604 discharges who had TAVR, 24.0% (5,674 patients) had had a diagnosis of SHF. The mean age of the overall TAVR cohort was 80.5 (standard deviation [SD] 8.3) years, 45.9% female. History of coronary artery disease, hypertension, hyperlipidaemia, and diabetes were the most common comorbidities. More than 96% of the TAVR were endovascular approach. After propensity matching, both TAVR-SHF and TAVR-C groups were comparable in terms of baseline characteristics (table 1).
Table 1. Demographics, baseline characteristics and comorbidities of overall all transcatheter aortic valve replacement (TAVR) group, and TAVR-systolic heart failure (TAVR-SHF) and control (TAVR-C) groups after propensity matching with p values for each variable
| Baseline/group | Overall TAVR* | TAVR-SHF | TAVR-C | p value** |
|---|---|---|---|---|
| Number of patients | 23,604 | 5,674 (24%) | 5,674 (24%) | – |
| Mean age, years (SD) | 80.5 (8.3) | 79.7 (8.9) | 80.1 (8.5) | <0.01 |
| Female | 45.9% | 35.5% | 35.7% | 0.72 |
| Transapical TAVR | 3.4% | 3.6% | 3.4% | 0.50 |
| Endovascular TAVR | 96.6% | 96.5% | 96.7% | 0.53 |
| Hypertension | 86.7% | 86.0% | 87.1% | 0.09 |
| Diabetes mellitus | 36.1% | 38.6% | 37.7% | 0.27 |
| Hyperlipidaemia | 69.0% | 67.9% | 68.3% | 0.66 |
| Chronic kidney disease | 34.3% | 43.1% | 42.7% | 0.65 |
| Coronary artery disease | 70.5% | 77.9% | 78.4% | 0.41 |
| Mitral regurgitation | 6.9% | 9.3% | 9.3% | 0.87 |
| Chronic anaemia | 23.4% | 25.7% | 24.9% | 0.31 |
| Atrial fibrillation | 40.2% | 47.6% | 46.9% | 0.45 |
| Atrial flutter | 4.1% | 5.5% | 5.0% | 0.20 |
| Long-term anticoagulation | 18.6% | 21.8% | 21.9% | 0.83 |
| Long-term antiplatelet | 10.1% | 9.9% | 10.0% | 0.77 |
| Aspirin | 27.9% | 26.9% | 27.2% | 0.69 |
| Abnormal coagulation profile | 0.4% | 0.6% | 0.3% | 0.02 |
| Peripheral vascular disease | 25.7% | 27.1% | 26.3% | 0.33 |
| Chronic pulmonary disease | 30.0% | 32.8% | 32.8% | 0.93 |
| Chronic liver disease | 3.0% | 2.8% | 2.6% | 0.44 |
| Smoking | 37.7% | 38.2% | 38.4% | 0.82 |
| Obesity | 17.1% | 15.8% | 15.0% | 0.21 |
| A p value <0.05 indicates a statistically significant difference between TAVR-SHF and TAVR-C groups. *Overall TAVR cohort characteristics before propensity matching. **p value evaluating the statistical difference between the TACR-SHF and TAVR-C after propensity matching. Key: SD = standard deviation |
||||
Table 2 shows, in comparison with TAVR-C, TAVR-SHF was associated with significantly higher in-hospital all-cause mortality (2.7% vs. 1.9%, p<0.01), two days longer hospital stay (7.5 vs. 5.5 days, p<0.01), higher cardiogenic shock (5.1% vs. 1.6%, p<0.01), AMI (4.0% vs. 1.9%, p<0.01), AKI (18.7% vs. 12.4%, p<0.01) and mechanical complications of prosthetic valve (1.2% vs. 0.6%, p<0.01). There was no significant difference between TAVR-SHF and TAVR-C in terms of bleeding (19.5% vs. 18.2%, p=0.08) and 30-day readmission rate (10.8% vs. 10.2%, p=0.29). Multi-variable analysis showed that hypertension, hyperlipidaemia, smoking, and long-term anticoagulation were predictors of increased mortality.
Table 1. Patient information and electrocardiogram (ECG) measurements
| End point/group | Overall TAVR* | TAVR-C | p value** |
|---|---|---|---|
| All-cause mortality | 1.8% | 1.9% | <0.01 |
| Mean LOS (SD) | 5.6 (7.3) days | 5.5 (6.9) days | <0.01 |
| Cardiogenic shock | 2.4% | 1.6% | <0.01 |
| Mechanical complication of prosthetic valve | 0.7% | 0.6% | <0.01 |
| AMI | 2.3% | 1.9% | <0.01 |
| AKI | 12.0% | 12.4% | <0.01 |
| Bleeding | 18.4% | 18.2% | 0.08 |
| Bleeding | 18.4% | 18.2% | 0.08 |
| Bleeding | 18.4% | 18.2% | 0.08 |
| Bleeding | 18.4% | 18.2% | 0.08 |
| Bleeding | 18.4% | 18.2% | 0.08 |
| End point/group | Overall TAVR* | TAVR-C | p value** |
| Bleeding | 18.4% | 18.2% | 0.08 |
| 30-day readmission | 11.6% | 10.2% | 0.29 |
| A p value <0.05 indicates a statistically significant difference between TAVR-SHF and TAVR-C groups. *Overall TAVR cohort characteristics before propensity matching. **p value evaluating the statistical difference between the TACR-CTO and TAVR-C after propensity matching. Key: AKI = acute kidney injury; AMI = acute myocardial infarction; LOS = length of stay; PPM = need for a new permanent pacemaker; SD = standard deviation |
|||
Discussion
This study shows that TAVR-SHF was associated with a higher risk of in-hospital all-cause mortality and comorbidity, including development of post-procedural cardiogenic shock, and longer LOS in comparison with the overall TAVR group and with TAVR-C after propensity matching adjustment to baseline comorbidities. There were no significant differences in terms of 30-day readmission rates.
Severe AS patients undergoing TAVR, especially those with systolic dysfunction, have limited myocardial reserve and less tolerance to acute haemodynamic changes, which can lead to cardiogenic shock.15,16 Intra-procedural circulatory depression occurs in up to 20%, which may worsen coronary perfusion and start a vicious cycle of ischaemia and systolic dysfunction.16,17 It is the most common cause of early (i.e. within 30 days) post-TAVR death on autopsy studies.18 There are multiple reasons for post-TAVR cardiogenic shock including mechanical complications of the valve. Furthermore, SHF represents an independent risk factor for mechanical complications of TAVR valve, consistent with this study result.16,17,19 The increased cardiogenic shock in TAVR-SHF possibly contributed to the higher in-hospital mortality, morbidity, and subsequently longer LOS.
SHF in severe AS can be primarily driven by the increase in afterload, which has a high probability of recovery after valve replacement. It can also be secondary to other causes, such as myocardial ischaemia, which leads to scar-induced SHF where recovery is less certain and prognosis might not substantially improve with TAVR.5 These different mechanisms of SHF in TAVR patients might be responsible for the conflicting results in the literature.
Similar readmission rates may be attributed to excellent procedural success, systolic function recovery, reverse remodelling, improved haemodynamics and functional status in patients who survived to discharge, regardless of baseline systolic function, as most of the late post-TAVR mortality is secondary to non-cardiovascular causes.8,18,20 The similar bleeding rates were probably because of utilisation of similar procedural techniques and approaches.
This study has important clinical implications. It highlights a large high-risk subgroup of TAVR patients who are at higher risk of post-procedural mortality and morbidity, who require careful peri-procedural observation and familiarity with potential complications. It raises the question of whether routine pre-TAVR evaluation (such as myocardial strain and stroke volume index) is needed to assess which patients with SHF would benefit the most with respect to ventricular function recovery. Furthermore, it raises a question regarding the role of prophylactic mechanical circulatory support devices during TAVR in select SHF patients who are at higher risk for cardiogenic shock.
Limitations
This study is a retrospective study. The NRD database does not provide information about the symptomatic class, severity (some data have shown that emergent TAVR for decompensated severe AS was feasible but associated with higher complications and mortality rates),21,22 cause of SHF, the type of TAVR valve, echocardiographic haemodynamics and biomarkers of AS and SHF that all might affect post-TAVR outcomes and prognosis, as ejection fraction by itself might not completely reflect extent of myocardial damage and is not as sensitive as myocardial strain.5,23 The post-TAVR ejection fraction, functional status, and long-term outcomes could not be assessed.
Conclusion
The presence of SHF is associated with increased early post-TAVR mortality and morbidity. Further research is warranted to prospectively delineate the peri-procedural risks by using newer echocardiographic parameters and role of using prophylactic mechanical support devices.
Key messages
- Systolic heart failure (SHF) is a common comorbidity in patients undergoing transcatheter aortic valve replacement (TAVR)
- The effect of SHF on the TAVR outcomes is unclear
- The presence of SHF is associated with increased early morbidity and mortality
Conflicts of interest
None declared.
Funding
None.
Study approval and consent
No ethical approval or consent was required.
References
1. Villari B, Vassalli G, Betocchi S, Briguori C, Chiariello M, Hess OM. Normalization of left ventricular nonuniformity late after valve replacement for aortic stenosis. Am J Cardiol 1996;78:66–71. https://doi.org/10.1016/S0002-9149(96)00229-9
2. Nishimura RA, Otto CM, Bonow RO et al. 2014 AHA/ACC guideline for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines. J Am Coll Cardiol 2014;63:e57–e185. https://doi.org/10.1161/CIR.0000000000000031
3. Elmariah S, Palacios IF, McAndrew T et al. Outcomes of transcatheter and surgical aortic valve replacement in high-risk patients with aortic stenosis and left ventricular dysfunction. Circ Cardiovasc Interv 2013;6:604–14. https://doi.org/10.1161/CIRCINTERVENTIONS.113.000650
4. Nishimura RA, Otto CM, Bonow RO et al. 2017 AHA/ACC focused update of the 2014 AHA/ACC guideline for the management of patients with valvular heart disease. J Am Coll Cardiol 2017;70:252–89. https://doi.org/10.1016/j.jacc.2017.03.011
5. Himbert D, Vahanian A. Transcatheter aortic valve replacement for patients with heart failure. Heart Fail Clin 2015;11:231–42. https://doi.org/10.1016/j.hfc.2014.12.003
6. Barili F, Pacini D, Capo A et al. Reliability of new scores in predicting perioperative mortality after isolated aortic valve surgery: a comparison with the Society of Thoracic Surgeons score and Logistic EuroSCORE. Ann Thorac Surg 2013;95:1539–44. https://doi.org/10.1016/j.athoracsur.2013.01.058
7. Silaschi M, Conradi L, Seiffert M et al. Predicting risk in transcatheter aortic valve implantation: comparative analysis of EuroSCORE II and established risk stratification tools. Thorac Cardiovasc Surg 2015;63:472–8. https://doi.org/10.1055/s-0034-1389107
8. Ewe SH, Marsan NA, Pepi M et al. Impact of severe left ventricular dysfunction on mid-term mortality in elderly patients undergoing transcatheter aortic valve implantation. Am Heart J 2010;160:1113–20. https://doi.org/10.1016/j.ahj.2010.09.003
9. Agency for Healthcare Research and Quality. 2014 Introduction to the NRD. Healthcare Cost and Utilization Project (HCUP). Rockville, MD, USA: Agency for Healthcare Research and Quality, December 2016. Available from: https://www.hcup-us.ahrq.gov/db/nation/nrd/NRD_Introduction_2014.jsp
10. Gibson T, Casto A, Young J, Karnell L CN. Impact of ICD-10-CM/PCS on research using administrative databases. HCUP Methods Series Report # 2016-02. Rockville, MD, USA: Agency for Healthcare Research and Quality, July 2016. Available from: https://www.hcup-us.ahrq.gov/reports/methods/2016-02.pdf
11. Khera R, Angraal S, Couch T et al. Adherence to methodological standards in research using the National Inpatient Sample. JAMA 2017;318:2011–18. https://doi.org/10.1001/jama.2017.17653
12. Yoon F, Sheng M, Jiang H, Steiner C, Barrett M. Calculating Nationwide Readmissions Database (NRD) variances. HCUP Methods Series Report # 2017-01. Rockville, MD, USA: Agency for Healthcare Research and Quality, January 2017. Available from: https://www.hcup-us.ahrq.gov/reports/methods/2017-01.pdf
13. Pourhoseingholi MA, Baghestani AR, Vahedi M. How to control confounding effects by statistical analysis. Gastroenterol Hepatol Bed Bench 2012;5:79–83. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4017459/
14. McDonald JH. Handbook of Biological Statistics. 3rd ed. Baltimore, Maryland: Sparky House Publishing, 2014.
15. Baldasseroni S, Opasich C, Gorini M et al. Left bundle-branch block is associated with increased 1-year sudden and total mortality rate in 5517 outpatients with congestive heart failure: a report from the Italian network on congestive heart failure. Am Heart J 2002;143:398–405. https://doi.org/10.1067/mhj.2002.121264
16. Alraies MC, Soud M, Pacha HM, Waksman R, Ben-Do I. Cardiogenic shock complicating transcatheter aortic valve replacement due to severe para-valvular regurgitation. Cardiovasc Revascularization Med 2018;19:393–5. https://doi.org/10.1016/j.carrev.2017.11.002
17. Laborde J-C, Brecker SJD, Roy D, Jahangiri M. Complications at the time of transcatheter aortic valve implantation. Methodist Debakey Cardiovasc J 2012;8:38–41. https://doi.org/10.14797/mdcj-8-2-38
18. van Kesteren F, Wiegerinck EMA, Rizzo S et al. Autopsy after transcatheter aortic valve implantation. Virchows Arch 2017;470:331–9. https://doi.org/10.1007/s00428-017-2076-4
19. Wilczek K, Bujak K, Reguła R, Chodór P, Osadnik T. Risk factors for paravalvular leak after transcatheter aortic valve implantation. Kardiochir Torakochirurgia Pol 2015;12:89–94. https://doi.org/10.5114/kitp.2015.52848
20. Ferrante G, Presbitero P, Pagnotta P et al. Impact of severe left ventricular dysfunction on mid-term mortality in elderly patients undergoing transcatheter aortic valve implantation. J Geriatr Cardiol 2016;13:290–8. https://doi.org/10.11909/j.issn.1671-5411.2016.04.001
21. D’Ancona G, Pasic M, Buz S et al. Transapical transcatheter aortic valve replacement in patients with cardiogenic shock. Interact Cardiovasc Thorac Surg 2012;14:426–30. https://doi.org/10.1093/icvts/ivr095
22. Frerker C, Schewel J, Schlüter M et al. Emergency transcatheter aortic valve replacement in patients with cardiogenic shock due to acutely decompensated aortic stenosis. Eurointervention 2016;11:1530–6. https://doi.org/10.4244/EIJY15M03_03
23. Herrmann HC, Pibarot P, Hueter I et al. Predictors of mortality and outcomes of therapy in low-flow severe aortic stenosis: a placement of aortic transcatheter valves (PARTNER) trial analysis. Circulation 2013;127:2316–26. https://doi.org/10.1161/CIRCULATIONAHA.112.001290
