Aortic valve replacement

Aortic valve replacement is a procedure unique in its ability to resolve significant and worsening aortic valve disease of any aetiology. The valve substitute can be a mechanical valve (usually a bileaflet valve), bioprosthesis (stented, stentless, sutureless, or transcathetral), allograft, or autograft. Specific advantages and drawbacks of respective type of the valve substitute are discussed in detail. The in-hospital mortality of aortic valve replacement ranges between 2 and 3%. Potential valve-related complications comprise thromboembolic and bleeding events, valve thrombosis, degeneration of bioprostheses, paravalvar leak, and patient–prosthesis mismatch.


Introduction
Left ventricular dysfunction increased the risk of surgical aortic valve replacement (SAVR) in patients with severe aortic stenosis. 1 However, SAVR had survival advantages in those patients compared with medical therapy. 2 The effect of low ejection fraction (EF) on the outcomes after transcatheter aortic valve replacement (TAVR) is controversial. 3 TAVR offered comparable outcomes in patients with low and preserved left ventricular function. 4 There is a lack of studies comparing SAVR and TAVR in patients with low EF, and there is no consensus on the recommended approach in those patients. Therefore, the present study compared hospital and mid-term outcomes, including survival, aortic valve reintervention, stroke, and readmission for heart failure in patients with an EF of ≤40% who underwent SAVR vs TAVR.

Design and Patients
We conducted a retrospective study that included 187 patients who underwent SAVR (n = 70) or TAVR (n = 117) at Prince Sultan Cardiac Center, Riyadh, Saudi Arabia. Patients who had SAVR were operated upon from 2002 to 2019, and TAVR was performed from 2009 to 2019. We included patients who had isolated aortic valve intervention with an EF ≤40%. Patients who had concomitant cardiac surgery or transcatheter intervention were excluded from our study. Patient assignment to each group was based on the Heart Team discussion and was based on patient risk stratification according to Euro SCORE II. 5 The Ethical Committee of Prince Sultan Cardiac Center approved data collection for this study (Reference number R19021); the need for patient consent was waived because of the retrospective design.

Study Data and Outcomes
Patient data were retrieved from the electronic and paper charts and the prospectively maintained TAVR registry. We compared demographic data, comorbidities, and laboratory data between groups. Operative data included operative urgency, valve type, and size.
Hospital outcomes were stroke, renal impairment, myocardial infarction (MI), postoperative atrial fibrillation (AF), permanent pacemaker (PPM) insertion, bleeding, hospital stay, and all-cause mortality. Long-term outcomes were survival and the composite endpoint of aortic valve reintervention, readmission for heart failure, and stroke.

Follow-Up
Patients had postoperative echocardiographic follow-ups according to the discretion of the treating physicians. There were 538 echocardiographic follow-ups available for all patients over a 5 year period. The changes in EF were traced longitudinally and compared between both groups. Patients were contacted by phone in June 2021 to confirm their vital status.

Definitions
A reduced EF was defined as an EF ≤40%. 6,7 Hospital outcomes were defined as those occurring during hospital admission or within 30 days from the procedure. Postoperative myocardial infarction (MI) was diagnosed with a pathological Q-wave, left bundle branch block, or angiographic evidence of coronary occlusion. 8 Postoperative renal impairment was defined as a 1.5-time increase in serum creatinine. 9 Preoperative variables were collected according to Euro SCORE II definitions. 5

Statistical Analysis
Continuous data were presented as median and interquartile range or mean and standard deviation. The student t-test was used to compare normally distributed continuous data and the Wilcoxon test was used for non-normal data. Categorical data were presented as frequencies and percentages and compared with the Chi-square or Fisher exact test. Time to event data was plotted using a Kaplan-Meier curve and compared between groups with the log-rank test. Univariable competing risk regression was used to assess the effect of the intervention on the composite endpoint of stroke, aortic valve reintervention, and admission for heart failure in the presence of death as a competing factor. Risk factors for mortality were evaluated using multivariable Cox regression analysis. Univariable Cox regression was used, and variables with a P < .15 were entered into a stepwise multivariable regression with a backward selection with a stay P < .1. Factors included in the multivariable regression were age, gender, creatinine clearance, hypertension, diabetes mellitus, previous myocardial infarction, previous percutaneous coronary intervention, preoperative mitral stenosis, extracardiac arteriopathy, left ventricular end-diastolic and end-systolic diameters, indexed left ventricular mass and the intervention group. Collinearity was tested with variance inflation factor (VIF), and all variables had a VIF <1.5. Random effect regression was used to compare the change in the EF between groups. Stata 16 (Stata Corp. College Station, TX, USA) was used for all analyses.

Baseline Data
The patients who had TAVR were significantly older, with significantly more females. Diabetes, hypertension, previous myocardial infarction (MI), extracardiac arteriopathy, and heart failure were more common in TAVR patients. Creatinine clearance was significantly lower in TAVR patients (Table 1).
Aortic valve regurgitation was more prevalent in the SAVR group. Patients who underwent TAVR had a higher prevalence of aortic stenosis, concomitant mitral valve regurgitation and stenosis, and tricuspid valve regurgitation. Patients in the SAVR group had significantly higher left ventricular enddiastolic and end-systolic diameters and pulmonary artery systolic pressure (Table 1).

Hospital Outcomes
There were no differences in stroke, myocardial infarction, and hospital mortality between groups. Postoperative renal impairment and prolonged hospital stay were more common in SAVR patients. Grade I paravalvular leak was significantly higher in the TAVR group. Permanent pacemaker insertion was non-significantly higher in the TAVR group (Table 3). Three patients in the SAVR group required re-exploration for bleeding, and 25 patients in the TAVR group had minor

Long-Term Outcomes
The median follow-up was 30 (15-72) months [73.5 (27-135) months in the SAVR group and 26 (12-39) months in the TAVR group]. The composite endpoint of stroke, aortic valve reintervention, and readmission for heart failure occurred in 13 patients. Heart failure admission occurred in 2 patients in the SAVR group and 6 in the TAVR group, stroke occurred in 4 patients in the SAVR group and 4 in TAVR, and aortic valve reintervention was required in 4 patients in the SAVR group (redo SAVR = 3, valve-in-valve = 1) and no patients in the TAVR group.
Competing risk regression of the composite endpoint (stroke, aortic valve reintervention, and readmission for heart failure) and in the presence of death as a competing risk factor elucidated no difference between both groups (Subdistributional Hazard Ratio (SHR): .95 (95% CI: .37-2.45); P = .92). There was no difference between groups regarding the composite endpoint ( Figure 1).
Mortality occurred in 34 patients (8 in the SAVR group and 26 in the TAVR group). Survival at 1, 3, and 5 years was 95%, 93%, and 91% in the SAVR group and 89%, 80%, and 63% in the TAVR group (log-rank P < .001) (Figure 2). Predictors of mortality were low creatinine clearance, diabetes mellitus, previous PCI, recent heart failure, and preoperative mitral stenosis. There was no difference in survival between SAVR and TAVR groups in the multivariable regression analysis (Table 4).

Discussion
Low EF is associated with increased postoperative morbidity and mortality in cardiac surgery patients. 10,11 Additionally, low EF was found as a risk factor for 30-day mortality in patients undergoing aortic valve replacement for severe aortic stenosis. 12 Despite that, SAVR was associated with good longterm outcomes with the recovery of ventricular function and functional class. 13,14   TAVR was presented as an attractive alternative to SAVR, and the indications were expanded to include low-risk patients. 15 The benefits vs the risks of TAVR in patients with low EF are controversial. Low EF (˂40%) was a predictor of early mortality in patients who underwent TAVR in the Italian registry. 16 Urena and associates, in a study of 3726 TAVR patients, found that low EF (≤40%) was an independent predictor of death during a 2-year follow-up. 17 In contrast to these studies, the Placement of AoRTic TraNscathetER Valve Trial (PARTNER trial) did not elucidate a significant effect of low EF on TAVR or SAVR outcomes. 18 There is a lack of studies comparing SAVR and TAVR in patients with low EF. In this study, we compared 70 patients who had SAVR and 117 patients who had TAVR, and all patients had isolated procedures with an EF ≤40%. In our series, SAVR patients were significantly younger; however, the age of TAVR patients was comparable with other trials. 15 Comorbidities were more prevalent in TAVR patients and were of moderate and high risk according to Euro Score. 5 The valve size was significantly bigger with TAVR, similar to other studies. 19 SAVR had a more postoperative renal impairment and longer hospital stay, with no difference in stroke and hospital mortality between groups. In a meta-analysis comparing SAVR and TAVR, there were no differences in 30-day mortality, stroke, acute kidney injury, and myocardial infarction between groups. 20 We did not find a difference between groups in the composite endpoint of heart failure readmission, aortic valve reintervention, and stroke. Survival was better in SAVR; however, after adjusting baseline data, the aortic valve  replacement approach did not affect survival. In a metaanalysis of SAVR and TAVR in intermediate and low-risk patients, no difference in 30-day, 1-, 2-, and 3-years mortality between groups. 21 The authors also did not find a difference in stroke and aortic valve reintervention in a 3-year follow-up. In our study, EF improved significantly in both groups during follow-up, and the most noticeable improvement occurred during the first year after the intervention. Robiolio and colleagues found a significant improvement in EF after SAVR in a 6-months follow-up. 22 Recovery of left ventricular function was observed to occur for many years after SAVR. 23   Similar findings were observed after TAVR with immediate improvement of EF, especially in patients with low EF. 24 Moreover, TAVR improved EF in low flow and low gradient patients with impaired ventricular function. 4 A budget impact analysis was conducted in Saudi Arabia and included our institution comparing TAVR vs SAVR. The study reported cost savings with TAVR. However, the study reported the 5-year cost only, and it was not limited to patients with low EF. 25 The major limitations of the study are the retrospective design and the lack of randomization. Patients were assigned to each group based on their characteristics, which led to significant variations in the baseline characteristics. The study outcomes could have been affected by patient characteristics. However, we adjusted the patient characteristics and found no difference in survival. On the other hand, other events had low frequency, and multivariable analysis was not feasible.
TAVR could be considered as safe as SAVR in patients with low EF. The EF improved significantly and similarly with both approaches.

Author Contributions
All authors contributed to: (1) substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data, (2) drafting the article or revising it critically for important intellectual content, and, (3) final approval of the version to be published.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.