Abstract
Background
Reduced leaflet motion (RLM) of transcatheter aortic valves (TAV) is observed in up to 4% of cases with similar frequency in surgical valves, with an overall incidence that differs based on prosthesis type and size. This phenomenon likely represents subclinical leaflet thrombosis. Herein we sought to analyze the existing reported literature to assess whether or not RLM is associated with subsequent valve degeneration or cerebrovascular events.
Methods and results
We searched PubMed, and EMBASE (2008–2017) to identify relevant studies. Studies with <1-year follow-up, studies not evaluating RLM, and/or clinical outcomes were excluded. Our co-primary endpoints were the incidence of cerebrovascular events (stroke and/or transient ischemic attack-TIA) or structural valvular degeneration defined as moderate or greater regurgitation and/or a mean gradient ≥20 mm Hg. The literature search yielded 30 potential studies. Of these, six observational studies with a total population of 1704 patients met our selection criteria. RLM was associated with an increased risk of stroke or TIA (adjusted OR 2.60, 95% CI 1.56 to 4.34, p = 0.004). At one year, RLM was associated with an increased risk of structural valve degeneration (adjusted OR 2.51, 95% CI 1.47 to 4.30, p = 0.006). The association between RLM and clinical endpoints remained even after limiting analysis to transcatheter aortic valve replacement (TAVR) patients only.
Conclusions
In patients with bio prosthetic aortic valve, presence of RLM is associated with increased risk of stroke or TIA as well as structural valvular degeneration. These findings support ongoing surveillance efforts and evaluation of pharmacotherapies to address RLM in effort to minimize subsequent clinical events.
Highlights
- •
Reduced Leaflet motion (RLM) of transcatheter aortic valves (TAV) is observed in up to 4% of cases with similar frequency in surgical valves. Herein we sought to analyze the existing reported literature to assess whether or not RLM is associated with subsequent valve degeneration or cerebrovascular events.
- •
Our co-primary endpoints were the incidence of cerebrovascular events (stroke and/or transient ischemic attacks-TIA) or structural valvular degeneration defined as moderate or greater regurgitation and/or a mean gradient > 20 mm Hg.
- •
Out of 30 studies, six observational studies with a total population of 1,704 patients met our selection criteria. RLM was associated with an increased risk of stroke or TIA (adjusted OR 2.60, 95% CI 1.56 to 4.34, p = 0.004). At one year, RLM was associated with an increased risk of structural valve degeneration (adjusted OR 2.51, 95% CI 1.47 to 4.30, p = 0.006). The association between RLM and clinical endpoints remained even after limiting analysis to transcatheter aortic valve (TAVR) patients only.
- •
Our analysis demonstrates the following findings: (1) RLM post valve implantation is associated with an increased occurrence of stroke and/or TIA; (2) RLM was associated with an increased incidence of structural valve degeneration and (3) The association between RLM and clinical endpoints was unchanged when limiting analysis to TAVR patients only.
- •
It is reasonable to surmise that thromboembolism from the affected valve may underlie cerebrovascular events, and that valve degeneration may be preceded by thrombotic RLM prior to realization of an increased gradient. Consistent with this, valve degeneration in the studies we analyzed was primarily related to an increase in gradient rather than valvular regurgitation.
1
Introduction
Transcatheter aortic valve replacement (TAVR) yields outcomes comparable to surgical replacement for patients with symptomatic severe aortic stenosis at intermediate or higher surgical risk [ ]. Recently post-TAVR computed tomography (CT) identified a subset of patients with reduced leaflet motion (RLM) [ ]. This observation led to the examination of leaflet motion across all transcatheter valve types, as well comparisons to historical and contemporary surgical valves [ ]. RLM has been observed with similar frequencies in both TAVR (1–4%) and surgical aortic valve replacement-SAVR (1–2%) procedures with an incidence that differs based on prosthesis type and size. Other variables such as implant depth, sinus of Valsalva size, and post-procedure anticoagulation strategy are also correlated with RLM [ ].
The etiology of RLM is theorized to be subclinical valve thrombosis associated with coincident leaflet motion abnormalities. RLM has a lower incidence in those on anticoagulation and is associated with elevated levels of coagulation biomarkers (D-dimer), and usually improves with therapeutic anticoagulation [ , , ]. Herein we sought to analyze the existing reported literature to assess whether or not RLM is associated with subsequent adverse clinical incidents, namely valve degeneration or cerebrovascular events in patients with both surgical and transcatheter aortic valves.
2
Methods
Our analysis is based on the guidelines of the Meta-analysis of Observational Studies in the Epidemiology Group [ ]. Briefly, these guidelines proposed specifications for reporting of meta-analyses of observational studies, including background, search strategy, methods, results, discussion, and conclusion. It was noted that use of the checklist improved the usefulness of meta-analyses for authors, reviewers, editors, readers, and decision makers [ ].
2.1
Inclusion criteria and search strategies
We included both prospective and retrospective observational studies with the primary objective to analyze the association between hypoattenuated leaflet thickening (HALT) and two clinical outcomes: (1) stroke or transient ischemic attack (TIA); and (2) structural valve degeneration. We searched PubMed, and EMBASE (2000–2017) to identify relevant studies. Titles and abstracts with the following terms were collated: ‘transcatheter aortic valve replacement’, ‘surgical aortic valve’, ‘bioprosthetic aortic valve’, ‘subclinical leaflet thrombosis’, ‘hypoattenuated leaflet thickening’, ‘reduced leaflet motion’, ‘TAVR thrombosis’, ‘TAVR HALT’, ‘SAVR thrombosis’, ‘SAVR HALT’, ‘stroke or all-stroke’, and ‘transient ischemic attack’. In addition, the ‘Related Articles’ feature on PubMed was used and a manual search was conducted using bibliographies of review articles on this topic. Abstracts of the articles published by the American College of Cardiology, the American Heart Association, the European Society of Cardiology, and Transcatheter Cardiovascular Therapies were also searched. Titles and abstracts were evaluated independently by two reviewers (N.M. and S.L.). Differences were resolved by consensus. These studies were then subjected to the following exclusion criteria: (i) no control group; (ii) no evaluation of leaflet thickening/reduced motion; (iii) a primary outcome different than stroke or TIA or valve degeneration; (iv) publication only in the abstract form; and (v) follow-up duration that is <1 year. Where multiple studies were done on the same cohort, we prioritized the most recent articles with the largest cohort size. Fig. 1 summarizes the results of the search strategy.
2.2
Quality assessment and data extraction
Studies were rated according to the Newcastle Ottawa Scale used for assessing nonrandomized observational studies [ ]. Two reviewers (N.M. and S.L.) extracted the following data elements: (i) publication details including first author’s last name, year; (ii) study design; (iii) characteristics of the study population including gender, race, mean age, comorbidities including hypertension, diabetes; (iv) variables included in the multivariate analyses; and (v) adjusted hazard ratio (HR) or odds ratio (OR) with 95% confidence interval (CI) from the multivariate analyses.
2.3
Clinical endpoints and definitions
Out study had two primary endpoints: 1) incidence of cerebrovascular events defined as stroke and or TIA; 2) incidence of structural valve degeneration.
Beginning with the earliest observations of RLM, terminology has expanded to encompass the likely pathophysiology of RLM – beginning with HALT, and if progressive, leading to RLM and perhaps other clinical sequelae. The definition of HALT includes “cusp thickness > 2 mm”, is associated with an elevated D-dimer, and represents the earliest image-based representation of leaflet thrombosis. It occurs independently of an increase in valve gradient (“subclinical leaflet thrombosis”), and precedes changes in leaflet excursion [ , ].
Stroke was defined as an episode of neurological dysfunction caused by focal cerebral, spinal, or retinal infarction. TIA is defined as an episode of neurological dysfunction with absence of positive imaging and resolution of symptoms within 24 h from onset [ ].
In regards to structural valve degeneration, all studies employed a uniform definition: moderate or greater regurgitation and/or a mean gradient ≥20 mm Hg.
2.4
Statistical analysis
The degree of association between RLM and stroke/TIA in patients with bioprosthetic aortic valve (SAVR and TAVR) was represented in terms of (OR). Summary OR and 95% confidence intervals (CIs) were calculated for all clinical outcomes by pooling published results available for each study. For all studies, multivariate regression analysis was performed to adjust for potential confounders (age, gender, history of atrial fibrillation, hypertension, diabetes, peripheral arterial disease, prior stroke/TIA, coronary artery disease and prior myocardial infarction). Calculated ORs were transformed logarithmically. We assessed heterogeneity of the studies by calculating a Q statistic (significance defined as p < 0.05), which we compared with the I 2 index (I 2 ≥ 56% defined as significant) [ , ]. Data were collected and analyzed using a random- and fixed-effect model approach with inverse-variance weighting [ ]. The underlying heterogeneity further prompted us to perform meta-regression analysis to investigate if our study end points (stroke/TIA and structural valvular degeneration) were affected by factors other than our primary risk factor (RLM). We adopted a weighted regression random-effect model and estimated between study variance (s 2 ) using empirical Bayes estimate. A two-sided p value < 0.05 was regarded as significant for all analyses. All statistical calculations were performed using RevMan v5.0 software (The Cochrane Collaboration, The Nordic Cochrane Centre, Copenhagen). Potential publication bias was represented graphically with Begg funnel plots of the natural log of OR vs. its standard error.
2
Methods
Our analysis is based on the guidelines of the Meta-analysis of Observational Studies in the Epidemiology Group [ ]. Briefly, these guidelines proposed specifications for reporting of meta-analyses of observational studies, including background, search strategy, methods, results, discussion, and conclusion. It was noted that use of the checklist improved the usefulness of meta-analyses for authors, reviewers, editors, readers, and decision makers [ ].
2.1
Inclusion criteria and search strategies
We included both prospective and retrospective observational studies with the primary objective to analyze the association between hypoattenuated leaflet thickening (HALT) and two clinical outcomes: (1) stroke or transient ischemic attack (TIA); and (2) structural valve degeneration. We searched PubMed, and EMBASE (2000–2017) to identify relevant studies. Titles and abstracts with the following terms were collated: ‘transcatheter aortic valve replacement’, ‘surgical aortic valve’, ‘bioprosthetic aortic valve’, ‘subclinical leaflet thrombosis’, ‘hypoattenuated leaflet thickening’, ‘reduced leaflet motion’, ‘TAVR thrombosis’, ‘TAVR HALT’, ‘SAVR thrombosis’, ‘SAVR HALT’, ‘stroke or all-stroke’, and ‘transient ischemic attack’. In addition, the ‘Related Articles’ feature on PubMed was used and a manual search was conducted using bibliographies of review articles on this topic. Abstracts of the articles published by the American College of Cardiology, the American Heart Association, the European Society of Cardiology, and Transcatheter Cardiovascular Therapies were also searched. Titles and abstracts were evaluated independently by two reviewers (N.M. and S.L.). Differences were resolved by consensus. These studies were then subjected to the following exclusion criteria: (i) no control group; (ii) no evaluation of leaflet thickening/reduced motion; (iii) a primary outcome different than stroke or TIA or valve degeneration; (iv) publication only in the abstract form; and (v) follow-up duration that is <1 year. Where multiple studies were done on the same cohort, we prioritized the most recent articles with the largest cohort size. Fig. 1 summarizes the results of the search strategy.
2.2
Quality assessment and data extraction
Studies were rated according to the Newcastle Ottawa Scale used for assessing nonrandomized observational studies [ ]. Two reviewers (N.M. and S.L.) extracted the following data elements: (i) publication details including first author’s last name, year; (ii) study design; (iii) characteristics of the study population including gender, race, mean age, comorbidities including hypertension, diabetes; (iv) variables included in the multivariate analyses; and (v) adjusted hazard ratio (HR) or odds ratio (OR) with 95% confidence interval (CI) from the multivariate analyses.
2.3
Clinical endpoints and definitions
Out study had two primary endpoints: 1) incidence of cerebrovascular events defined as stroke and or TIA; 2) incidence of structural valve degeneration.
Beginning with the earliest observations of RLM, terminology has expanded to encompass the likely pathophysiology of RLM – beginning with HALT, and if progressive, leading to RLM and perhaps other clinical sequelae. The definition of HALT includes “cusp thickness > 2 mm”, is associated with an elevated D-dimer, and represents the earliest image-based representation of leaflet thrombosis. It occurs independently of an increase in valve gradient (“subclinical leaflet thrombosis”), and precedes changes in leaflet excursion [ , ].
Stroke was defined as an episode of neurological dysfunction caused by focal cerebral, spinal, or retinal infarction. TIA is defined as an episode of neurological dysfunction with absence of positive imaging and resolution of symptoms within 24 h from onset [ ].
In regards to structural valve degeneration, all studies employed a uniform definition: moderate or greater regurgitation and/or a mean gradient ≥20 mm Hg.
2.4
Statistical analysis
The degree of association between RLM and stroke/TIA in patients with bioprosthetic aortic valve (SAVR and TAVR) was represented in terms of (OR). Summary OR and 95% confidence intervals (CIs) were calculated for all clinical outcomes by pooling published results available for each study. For all studies, multivariate regression analysis was performed to adjust for potential confounders (age, gender, history of atrial fibrillation, hypertension, diabetes, peripheral arterial disease, prior stroke/TIA, coronary artery disease and prior myocardial infarction). Calculated ORs were transformed logarithmically. We assessed heterogeneity of the studies by calculating a Q statistic (significance defined as p < 0.05), which we compared with the I 2 index (I 2 ≥ 56% defined as significant) [ , ]. Data were collected and analyzed using a random- and fixed-effect model approach with inverse-variance weighting [ ]. The underlying heterogeneity further prompted us to perform meta-regression analysis to investigate if our study end points (stroke/TIA and structural valvular degeneration) were affected by factors other than our primary risk factor (RLM). We adopted a weighted regression random-effect model and estimated between study variance (s 2 ) using empirical Bayes estimate. A two-sided p value < 0.05 was regarded as significant for all analyses. All statistical calculations were performed using RevMan v5.0 software (The Cochrane Collaboration, The Nordic Cochrane Centre, Copenhagen). Potential publication bias was represented graphically with Begg funnel plots of the natural log of OR vs. its standard error.
3
Results
The literature search yielded 30 potential studies. Of these, a total of six observational studies [ , , , , , ] with a total population of 1704 patients met our selection criteria ( Table 1 ). Eligible studies included combined data from the Assessment of Transcatheter and Surgical Aortic Bioprosthetic Valve Thrombosis and its Treatment with Anticoagulation (RESOLVE) and the Subclinical Aortic Valve Thrombosis Assessed with Four Dimensional Computed Tomography (SAVORY) registries. Four other studies were single center registries, while one specifically reported data from the Portico Re-sheathable Transcatheter Aortic Valve System U.S. Investigational Device Exemption (PORTICO IDE) study.
