Sutureless aortic valve replacement versus transcatheter
aortic valve implantation: a meta-analysis of comparative
matched studies using propensity
... [Show More] score matching
Massimo Mecoa,†
, Antonio Micelib,c,†
, Andrea Montiscib,*
,†
, Francesco Donatellib,d, Silvia Cirrib
,
Matteo Ferrarinib
, Antonio Liob
and Mattia Glauberb
a
Cardiac Centre, Humanitas Gavazzeni Hospital, Bergamo, Italy
b
Cardiothoracic Center, Istituto Clinico Sant’Ambrogio, Gruppo Ospedaliero San Donato, Milan, Italy
c Department of Clinical Science at South Bristol, University of Bristol, Bristol, UK
d Department of Cardiac Surgery, University of Milan, Milan, Italy
* Corresponding author. Cardiothoracic Centre, Istituto Clinico Sant’Ambrogio, Gruppo Ospedaliero San Donato, Via G. Faravelli 16, 20149 Milan, Italy.
Tel: +39-02-33127730; fax: +39-02-3312; e-mail: [email protected] (A. Montisci).
Received 25 March 2017; received in revised form 19 July 2017; accepted 30 July 2017
Abstract
OBJECTIVES: The aim of this meta-analysis was to compare outcomes of patients undergoing transcatheter aortic valve implantation
(TAVI) with those undergoing surgical aortic valve replacement using sutureless valves.
METHODS: A systematic review and meta-analysis in accordance with the Preferred Reporting Items for Systematic Reviews and MetaAnalyses (PRISMA) statement was performed.
RESULTS: No randomized controlled trials were identified. Six comparative studies using propensity score matching met the inclusion criteria. This meta-analysis identified 1462 patients in that 731 patients underwent surgical aortic valve replacement using sutureless valves (SU)
and 731 patients underwent a TAVI. The 30-day or in-hospital mortality was lower in the SU group [odds ratio (OR) 0.54, 95% confidence
interval (CI) 0.36–0.80; P = 0.003]. In the TAVI group, the incidence of postoperative stroke was higher (OR 0.36, 95% CI 0.17–0.79; P = 0.01).
The incidence of moderate or severe paravalvular regurgitation was higher in the TAVI group (OR 0.22, 95% CI 0.14–0.35; P = 0.001). There
were neither differences in the postoperative renal failure (OR 1.44, 95% CI 0.46–4.58; P = 0.53) nor in the number of patients requiring postoperative pacemaker implantation (OR 1.06, 95% CI 0.54–2.08; P = 0.86). Patients in the SU group required more transfusions (OR 4.47, 95%
CI 2.77–7.21; P = 0.0001), whereas those in the TAVI group had higher major vascular complications (OR 0.06, 95% CI 0.01–0.25; P = 0.0001).
Intensive care unit stay was not different (mean difference 0.99, 95% CI - 1.22 to 1.40; P = 0.53). One-year survival was better in the SU group
(Peto OR 0.35, 95% CI 0.18–0.67; P = 0.001), as was the 2-year survival (Peto OR 0.38, 95% CI 0.17–0.86; P = 0.001).
CONCLUSIONS: Surgical aortic valve replacement using sutureless valves is associated with better early and mid-term outcomes compared with TAVI in high- or intermediate-risk patients.
Keywords: Aortic valve surgery • Sutureless bioprosthesis • Transcatheter aortic valve implantation • Minimally invasive cardiac surgery •
Meta-analysis
INTRODUCTION
Severe aortic stenosis is the most common valvular heart disease
with a prevalence of 2.8% in adults older than 75 years. Aortic valve
replacement (AVR) using a biological valve prosthesis is the conventional treatment option for severe aortic stenosis. Nevertheless,
in recent years, other therapeutic strategies have been developed
with the aim of reducing mortality and morbidity as well as the
invasiveness of surgical procedure. Transcatheter aortic valve
implantation (TAVI) has been emerged as a valuable therapeutic
tool in high-risk patients or in those patients ineligible for a surgical
operation [1]. Nevertheless, sutureless or rapid-deployment aortic
valves in combination with a minimally invasive approach have
shown excellent outcomes, allowing a reduction of myocardial
ischaemia and cardiopulmonary bypass times compared with conventional surgery [2]. Several studies have reported controversial
results when these 2 treatment options were compared in highand intermediate-risk patients. However, no randomized controlled trials (RCTs) have been designed. Therefore, we conducted
a meta-analysis of statistically sound studies to evaluate surgical
outcomes of patients undergoing TAVI versus surgical aortic valve
†The first three authors contributed equally to this work. replacement (SAVR) using sutureless valves (SU-AVR).
VC The Author 2017. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
Interactive CardioVascular and Thoracic Surgery 26 (2018) 202–209 ORIGINAL ARTICLE
doi:10.1093/icvts/ivx294 Advance Access publication 11 September 2017
Downloaded from https://academic.oup.com/icvts/article/26/2/202/4111235 by guest on 18 October 2021
MATERIALS AND METHODS
Data sources and search strategy
We performed a systematic review and a meta-analysis in
accordance with the standards set forth by the Preferred
Reporting Items for Systematic Reviews and Meta-Analyses
(PRISMA) statement. We searched the PubMed, the EMBASE and
the Cochrane Collaboration databases using the keyword
‘sutureless’ AND ‘aortic valve’, ‘bioprosthesis’ AND ‘aortic valve’,
‘rapid deployment’ AND ‘aortic valve’, ‘aortic valve’ AND ‘surgery’
AND ‘transcatheter valve replacement’ AND ‘transcatheter valve
implantation’ OR ‘operation’ OR ‘replacement’. In addition, we
hand searched the references of retrieved articles and used the
articles related to the PubMed database feature to identify studies not captured by our primary search strategy. The search was
limited to trials involving human subjects, and no language
restriction was imposed. The final search was run on 31
December 2016.
Study selection
We included only comparative studies that enrolled patients at
intermediate and high operative risk who were treated with
sutureless aortic valve or TAVI, with at least 1 matching technique
(propensity score-matched study). Once full articles were
retrieved, studies were further excluded if there was an overlap in
patients from other studies within the same analysis. In this case,
the largest sample size was selected. Thus, although some
patients could possibly have been included in both controlled
and uncontrolled study analyses, they were only included once
in any given analysis. Consequently, there was no overlap in
patients in our meta-analyses.
Date extraction, quality and risk-of-bias assessment
Two reviewers (M.M. and A.M.) evaluated each article separately.
Disagreements were resolved by discussion with a third party
(M.G.).
The quality of studies was examined using the method recommended by a Cochrane Collaboration tool for assessing risk of
bias in the included studies. We extracted data pertaining to
baseline characteristics of study subjects, trial inclusion and
exclusion criteria, postoperative complications, maximal length
of follow-up and mortality.
Outcomes
The primary end-point was 30-day mortality or in-hospital
all-cause mortality. Secondary end-points were the following:
(i) postoperative renal failure, (ii) postoperative stroke, (iii) postoperative transvalvular mean gradient, (iv) postoperative paravalvular leak (PVL), (v) postoperative pacemaker implantation,
(vi) 1-year mortality and (vii) 2-year mortality.
Statistical analysis
Meta-analysis of early outcomes. This meta-analysis was
conducted by searching for any differences in mortality between
the SU group and the TAVI group. The results from all of the
relative studies were combined to estimate the odds ratios (ORs)
and associated 95% confidence intervals (CIs) for dichotomous
outcomes. With respect to the continuous outcomes, weighted
mean differences and 95% CIs were estimated as the effect
results. The I
2
statistic was used to test for heterogeneity, and
studies were considered to have low (I
2
= 25–49%), moderate
(I
2
= 50–74%) or high (I
2
>_ 75%) heterogeneity. Thus, an I
2
>_ 50%
indicates significant heterogeneity. In such cases, a random-effect
model was used, whereas an inverse variance method of fixedeffect model was used when the outcome had no significant heterogeneity (I
2
< 50%). Whenever heterogeneity was present, we
performed sensitivity analyses to investigate the influence of a
single study by excluding 1 study in each turn. Publication bias
was assessed by funnel plot using mortality as an end-point.
A value of P < 0.05 was considered to be statistically significant in
this meta-analysis. Statistical analyses were performed using
Review Manager 5 software (RevMan 5, The Cochrane
Collaboration, Oxford, UK).
Meta-analysis of late outcomes. The Kaplan–Meier curves
were digitalized using a dedicated software (Plot Digitizer). Using
the software, the axes were defined, and the curve was read point
by point. The Kaplan–Meier data reconstruction was employed
to derive individual patient data. We measured the recurrence
rate, a type of time-to-event data, with the statistic hazard ratio
(HR) and its 95% CIs. Based on the article by Tierney et al. [7],
‘observed–expected events research (O-E)’ and ‘variance (V)’ were
generated from HRs obtained from various sources, which may
be presented as number of recurrence, a Kaplan–Meier curve or
others. We analysed the transformed data in RevMan 5.2 and
used the statistical method exp [(O-E)/V] to calculate the HR and
to implement the time-to-event data analysis with a fixed-effect
model.
Meta-regression analysis. To evaluate the effects of preoperative and postoperative parameters on outcome, we performed
mixed-effects (unrestricted maximum likelihood) meta-regression
analyses. The meta-regression coefficient (slope of meta-regression
line) shows the estimated increase in logarithmic OR per unit
increase in the covariate. Because logarithmic OR exceeding 0 corresponds to OR exceeding 1 and logarithmic OR of less than 0 corresponds to OR of less than 1, a negative coefficient would
indicate that as the given factor increases, the OR decreases. All
meta-regression analyses were conducted using Open Meta
Analyst software [3].
RESULTS
Characteristics of the included study
The study selection process is shown in Fig. 1. After exclusion of
duplicate or irrelevant references, 104 potentially relevant articles
were retrieved. After detailed evaluation, 6 comparative
‘statistically sound’ studies [4–10] met the inclusion criteria and
were included in this meta-analysis.
This meta-analysis identified 1462 patients in that 731 patients
underwent SU-AVR sutureless valve and 731 patients underwent a
TAVI procedure. A list of included studies with their characteristics
and quality assessment is shown in Table 1 and Supplementary
Material, Fig. S1.
ADULT CARDIAC
M. Meco et al. / Interactive CardioVascular and Thoracic Surgery 203
Downloaded from https://academic.oup.com/icvts/article/26/2/202/4111235 by guest on 18 October 2021
Patient characteristics
The preoperative characteristics are shown in Table 2. The preoperative risk was estimated using the logistic EuroSCORE I and
showed no differences between the 2 groups. All other analysed
preoperative variables were not different between the 2 groups
(Supplementary Material, Figs S2–S12).
Postoperative outcomes [Show Less]