The prognostic implications of several baseline preprocedural variables in patients with severe native valve aortic stenosis (AS) undergoing transcatheter aortic valve implantation (TAVI) are unclear. The goal of this study was to determine the impact of reduced stroke volume index (SVI), low gradient (LG), and reduced ejection fraction (EF) on all-cause mortality. We searched MEDLINE, CENTRAL, EMBASE, Web of Science, and Scopus through October 13, 2014. We evaluated the association between low SVI (<35 ml/m 2 ), LG (<40 mm Hg), and low EF (<50% and <30%) on 1-year all-cause mortality. We pooled results across studies using the random-effects model. We included 16 studies at moderate risk of bias enrolling 7,673 patients with severe AS who underwent TAVI. Low EF was associated with increased 1-year mortality after TAVI compared to preserved EF (for EF <30%, hazard ratio [HR] 1.60, 95% confidence interval [CI] 1.19 to 2.16, I 2 = 32%; and for EF <50%, HR 1.52, 95% CI 1.31 to 1.76, I 2 = 17%). LG was associated with increased mortality after TAVI compared to high mean gradient (≥40 mm Hg; HR 1.60, 95% CI 1.30 to 1.97, I 2 = 36%). Low SVI was associated with increased mortality after TAVI compared to normal SVI (HR 1.59, 95% CI 1.23 to 2.05, I 2 = 27%). In conclusion, low SVI, LG, and low EF are each associated with higher mortality after TAVI. These findings highlight the importance of including these variables into TAVI risk algorithms and will better inform shared decision-making before TAVI.
Recent investigations have suggested that reduced stroke volume index (SVI) may be an independent predictor of mortality after transcatheter aortic valve implantation (TAVI). In contrast, there have been conflicting data regarding the impact of low ejection fraction (EF) and low gradient (LG) on outcomes after TAVI. In addition there have been varying definitions used for low flow, LG aortic stenosis (AS) in the reports, with some studies using only EF for the definition and others not incorporating SVI into the determination, further contributing to confusion. Accordingly, the objective of this study was to evaluate the individual effects of reduced SVI, LG, and low EF on mortality in patients undergoing TAVI.
Methods
This systematic review is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines and followed an a priori established protocol. The quality of evidence was rated using the Grading of Recommendations Assessment, Development, and Evaluation approach.
We searched Ovid MEDLINE, Ovid CENTRAL, Ovid EMBASE, Web of Science, and Scopus from database inception to October 13, 2014, using a combination of controlled vocabulary (subject headings) and text words. All languages and all ages were included. The initial search was performed on Ovid MEDLINE In-Process and Other Non-Indexed citations and then modified to the other databases. Specific key words included transcatheter aortic valve implantation, TAVI, TAVR, aortic valve stenosis, low flow, low output, LG, low EF, stroke volume, and death. The detailed search strategy is available in the Supplementary Material 1 . In addition, the references of eligible articles and relevant citations were reviewed using Web of Science and Scopus.
Two different EF cutpoints were used to define low EF and used in separate analyses: (1) EF <50% and (2) EF <30%, with 1 study using a definition of EF <35% included in this category. LG was defined as a mean aortic valve gradient <40 mm Hg measured by Doppler echocardiography. Low SVI was defined as SVI measured by Doppler echocardiography <35 ml/m 2 .
In this systematic review, we included published studies meeting the following criteria: (1) population-based, single-center, or multicenter studies; (2) patients with native severe AS (defined as an aortic valve area <1 cm 2 ) undergoing TAVI; (3) reported all-cause mortality 1 year after TAVI. Both retrospective and prospective studies were included, provided complete (or near-complete [>75%]), minimum 1-year follow-up was available, to allow accurate estimation of survival. Studies of patients with both reduced (<50%) and/or preserved (≥50%) left ventricular EF were included in the analysis.
We excluded the following studies: (1) single case reports; (2) studies with inadequate information regarding EF, SVI, and/or mean transvalvular gradient; (3) studies in which outcome of interest was not reported or was not able to be calculated from the results even after contacting the principal investigator of the specific study. In case of multiple studies from the same cohort, we only included data from the most recent comprehensive report.
Citations were screened at the title and abstract level by 2 independent reviewers and retrieved as a full report if they reported data on outcomes after TAVI. Data on the following study- and patient-related characteristics were independently abstracted by 2 reviewers (and discrepancies resolved by consensus): (1) study characteristics—primary author, time period of study/year of publication, geographic location of the population studied, study design; (2) characteristics—total number of patients undergoing TAVI, reduced and/or preserved EF, patients stratified by LG and/or low SVI, type and access route of TAVI devices used; (3) risk factors (co-morbidities including previous coronary artery bypass grafting surgery, history of stroke, diabetes mellitus, chronic pulmonary disease, chronic kidney disease); and (4) outcomes—1-year all-cause or cardiovascular mortality. Corresponding authors of studies were contacted to provide additional information where required.
Two reviewers independently assessed the risk of bias of the included observational studies using the Newcastle-Ottawa instrument. This scale grades studies according to 8 methodologic criteria.
For each study, we estimated the risk ratio and the associated 95% confidence interval (CI). Risk ratios were pooled across studies using the random-effects model. Statistical heterogeneity was assessed using the I 2 test. We used the following arbitrary cutoffs of I 2 value of <40% to be considered low heterogeneity; 40% to 60%, moderate; and >60%, substantial heterogeneity. Statistical analysis was performed using Comprehensive Meta-Analysis Version 2 (Biostat, Inc., Englewood, New Jersey).
Results
A study selection process flowchart detailing the number of studies identified, evaluated, and excluded is shown in Figure 1 . A total of 16 studies were included in the final analysis, enrolling 7,673 patients ( Figure 1 ).
Sixteen studies were included in the EF analysis (n = 7673), 8 were included in the LG analysis (n = 3790), and 4 were included in the SVI analysis (n = 2032). Seven studies included outcomes for EF <30% and 10 included outcomes for EF <50%. Study quality characteristics according to the Newcastle-Ottawa scale are shown in Supplementary Material 2 . Most studies presented adjusted hazard ratios (HRs) for 1-year all-cause mortality (n = 11) or cardiovascular mortality (n = 1), whereas 3 studies presented unadjusted 1-year all-cause mortality. Most studies adjusted for age, gender, logistic EuroScore, and/or Society for Thoracic Surgeons score, whereas some adjusted for history of peripheral vascular disease, previous cerebrovascular event, previous coronary artery bypass grafting surgery, and postprocedural aortic regurgitation.
Most patient characteristics were similar across the included studies (mean age 79 to 84 years, Society for Thoracic Surgeons score 4.9 to 11.1, logistic EuroScore 16 to 30, and mean EF 51% to 56%; Table 1 ). Mean SVI varied widely among studies reporting this variable (27 to 52 ml/m 2 ). Among the 13 studies reporting 1-year all-cause mortality numbers (n = 6,216), there were 1,230 deaths at 1 year (20% rate).
| First Author | Year | Location | Study Design | THV Type | N | Age (Yrs) | Male | CAD | Prior CV Surgery | STS score | Logistic Euroscore | AVA (cm2) | EF (%) | MG (mmHg) | SVI (ml/m2) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Ewe | 2010 | Milan, Italy & Leiden, Netherlands | RC | ES | 147 | 80 | 43% | 35% | 19% | NA | 21.8 | 0.67 | 53 | 48 | NA |
| Moat | 2011 | Nationwide, United Kingdom | RC | Both | 870 | 82 | 52% | 45% | 30% | NA | 18.5 | NA | NA | NA | NA |
| Pilgrim | 2011 | Bern, Switzerland | RC | Both | 256 | 82 | 44% | 44% | 21% | 6.4 | 24.8 | 0.70 | 51 | 44 | NA |
| Gotzmann | 2012 | Bochum, Germany | RC | MC | 202 | 79 | 47% | 52% | 21% | NA | 22.0 | 0.70 | 53 | 47 | NA |
| Amabile | 2012 | LePlessis-Robinson, France | RC | Both | 126 | 83 | 41% | 60% | NA | NA | 21.1 | NA | 57 | 51 | 40 |
| Fraccaro | 2012 | Milan & Padova, Italy | RC | Both | 384 | 80 | 48% | 46% | 20% | 9.8 | 24.0 | 0.75 | 53 | 50 | NA |
| Van der Boom | 2012 | Rotterdam, Netherlands | RC | MC | 230 | 80 | 51% | 46% | 24% | 4.9 | 16.4 | 0.66 | 51 | 44 | NA |
| Herrmann | 2013 | USA | RCT | ES | 971 | 84 | 53% | 76% | 41% | 11.1 | 26.4 | 0.64 | 53 | 44 | 35 |
| Le Ven | 2013 | Quebec City & Vancouver, Canada | RC | ES | 639 | 81 | 49% | 63% | 34% | 7 | 21.7 | 0.63 | 54 | 43 | 35 |
| Zahn | 2013 | , Germany | RC | Both | 1318 | 82 | 44% | 52% | 18% | NA | 20.3 | 0.68 | 53 | 50 | NA |
| O’Sullivan | 2013 | Bern | RC | Both | 354 | 83 | 43% | 59% | 14% | 7 | 23.6 | 0.52 | 53 | 45 | 27 |
| Munoz Garcia | 2013 | Spain, Portugal, Argentina, Chile, Venezuela | RC | MC | 1220 | 81 | 45% | 36% | 10% | NA | 17.8 | 0.62 | 56 | 52 | NA |
| Piazza | 2013 | Mulinational, Germany, Switzerland, Netherlands | RC | Both | 405 | 81 | 42% | 57% | NA | NA | 20.9 | NA | NA | NA | NA |
| Biner, | 2014 | Tel Aviv | RC | MC | 112 | 83 | 74% | 90% | NA | NA | 25.3 | 0.66 | 56 | 48 | 52 |
| Elhmidi | 2014 | Munich | RC | Both | 68 | 79 | 62% | 65% | 25% | 8.6 | 29.6 | 0.62 | NA | 41 | NA |
| Barbash | 2014 | Washington, DC | RC | Both | 371 | 84 | 49% | 55% | 34% | 10.5 | 29.0 | 0.63 | 53 | 49 | NA |
In most of the individual studies, low baseline EF was associated with a trend toward increased mortality at 1 year ( Figure 2 ). In the meta-analysis regardless of the EF cutpoint used (30% or 50%), low EF was associated with increased mortality after TAVI compared to preserved EF, with the following pooled estimates for 1-year mortality: EF <30%, HR 1.60, 95% CI 1.19 to 2.16, I 2 = 32%; and EF <50%, HR 1.52, 95% CI 1.31 to 1.76, I 2 = 17%; Figure 2 ). Statistical heterogeneity was greater for EF <30% compared to EF <50%.
Results among studies reporting effects of LG were concordant with all reports showing either a trend or a significant increase in 1-year mortality associated with LG. Meta-analysis showed that low mean gradient (<40 mm Hg) was associated with increased mortality after TAVI compared to high mean gradient (≥40 mm Hg; pooled estimate for 1-year mortality, HR 1.60, 95% CI 1.30 to 1.97, I 2 = 36%; Figure 3 ).
