Skip main navigation
×

Radiographic and Clinical Brain Infarcts in Cardiac and Diagnostic Procedures

A Systematic Review and Meta-Analysis
Originally publishedhttps://doi.org/10.1161/STROKEAHA.117.017541Stroke. 2017;48:2753–2759

Abstract

Background and Purpose—

The incidence of periprocedural brain infarcts varies among cardiovascular procedures. In a systematic review, we compared the ratio of radiographic brain infarcts (RBI) to strokes and transient ischemic attacks across cardiac and vascular procedures.

Methods—

We searched MEDLINE and 5 other databases for brain infarcts in aortic valve replacement, coronary artery bypass grafting, cardiac catheterization, and cerebral angiogram through September 2015. We followed the PRISMA (preferred reporting items for systematic reviews and meta-analyses) recommendations. We defined symptomatic rate ratio (RR) as ratio of stroke plus transient ischemic attack rate to RBI rate.

Results—

Twenty-nine studies involving 2124 subjects met the inclusion criteria. In meta-analysis of aortic valve replacements with 494 people, 69.4% (95% confidence interval (CI), 57.6%–81.4%) had RBIs, whereas 3.6% (95% CI, 2.0%–5.2%) had clinical events (RR, 0.08; 95% CI, 0.05–0.12). Coronary artery bypass grafting among 204 patients had 27.4% (95% CI, 6.0%–48.8%) RBIs and 2.4% (95% CI, 0.3%–4.5%) clinical events (RR, 0.11; 95% CI, 0.05–0.26). Cardiac catheterization among 833 people had 8.0% (95% CI, 4.1%–12.0%) RBIs, and 0.6% (95% CI, 0.1%–1.1%) had clinical events (RR, 0.16; 95% CI, 0.08–0.31). Cerebral angiogram among 593 people had 12.8% (95% CI, 6.6–19.0) RBIs and 0.6% (95% CI, 0%–13%) clinical events (RR, 0.10; 95% CI, 0.04–0.27). The RR of all procedures was 0.10 (95% CI, 0.07–0.13) without differences in the RRs across procedures (P=0.29).

Conclusions—

One of 10 people with periprocedural RBIs during cardiac surgeries and invasive vascular diagnostic procedures resulted in strokes or transient ischemic attacks, which may serve as a potential surrogate marker of procedural proficiency and perhaps as a predictor of risk for periprocedural strokes.

Introduction

Silent brain infarctions are clinically asymptomatic radiographic brain infarcts (RBIs) that are found in neuroimaging studies. Diffusion-weighted imaging (DWI) sequence has emerged as a gold standard for the detection of periprocedural RBIs because of its high sensitivity of detecting acute ischemic brain lesions.1,2 RBIs are common after routine invasive cerebrovascular and cardiovascular procedures. The frequency of periprocedural acute ischemic brain lesions on magnetic resonance imaging (MRI) is higher than clinically overt symptomatic brain infarcts, such as strokes and transient ischemic attacks (TIAs). Along with clinical cerebral infarcts, RBIs have been reported as a marker for the risk of certain surgeries or procedures. In a review on frequency of silent brain infarctions and symptomatic cerebral infarcts in common invasive procedures, such as carotid interventions, cerebral angiography, cardiac catheterization, and cardiac surgeries, including aortic valve replacement (AVR) and coronary artery bypass grafting (CABG), the study suggested periprocedural ischemic events might serve as a potential surrogate marker for optimizing invasive procedures.3

For MRI DWI sequence to be useful surrogate marker for procedural complications, a firm relationship between stroke complications and RBI should be made. We performed a systematic review and meta-analysis of studies of clinical and RBIs after surgical AVR/repair (SAVR), transcatheter AVR (TAVR), CABG, diagnostic cardiac catheterization (coronary angiogram), and diagnostic cerebral angiogram. We hypothesized that there is a constant ratio of clinical-to-radiographic infarct rates across the different diagnostic procedures and cardiac surgeries, which may serve as a potential surrogate marker of procedural safety and perhaps as a predictor of risk for periprocedural strokes.

Subjects and Methods

Search Strategy

All procedures used in this meta-analysis were consistent with PRISMA (preferred reporting items for systematic reviews and meta-analyses) guidelines (Table I in the online-only Data Supplement).4 We searched MEDLINE (PubMed) and 5 other databases—EMBASE, Cochrane Library, Web of Science, clinicaltrials.gov, and Scopus—for subject headings and text related to brain infarcts in for AVR, coronary artery bypass, cardiac catheterization, and cerebral angiogram from inception through September 2015. We used the following search terms with all possible combinations: MR imaging, silent infarct, asymptomatic infarct, cerebral infarct, new brain lesion, cerebral embolism, silent embolism, new brain infarct, CABG, carotid endarterectomy (CEA), CEA, carotid artery stenting, cardiac catheterization, cerebral angiogram, cerebral aneurysm coiling, open surgery/repair (SAVR), and transcatheter aortic valve implantation. The search was refined to include only cardiac surgeries and diagnostic procedures. The search strategy for PubMed is available in Appendix I in the online-only Data Supplement.

Inclusion Criteria

Inclusion criteria were applied following the PICOS (population, intervention, comparator, outcome, and study design) approach.4 An original study was eligible for our meta-analysis (1) if surgical AVR/repair (AVR), TAVR/transcatheter aortic valve implantation, CABG, diagnostic cardiac catheterization (coronary angiogram), or diagnostic cerebral angiogram were performed; (2) if brain MRI was performed systematically pre- and post-procedure along with a postprocedure clinical neurological examination; and (3) if new RBI and stroke incidence were reported after the procedures. We included study in any languages. Study designs included both retrospective and prospective observational studies and randomized controlled trials.

Exclusion Criteria

We excluded case reports, editorials, commentaries, meta-analysis articles, review articles, animal studies, and the articles with pediatric population (age <18). We excluded studies that reported infarct incidence in a common study population to avoid duplication and chose studies with larger and more inclusive studies. Exclusion of articles were discussed in detail among authors (S.-M.C. and K.U.).

Data Extraction and Management

Two reviewers (S.-M.C. and A.D.) assessed studies for inclusion and extracted information from articles, such as study design (study type and sample size), patient characteristics, and study procedures (type of interventions and method of stroke ascertainment), and study results (number of people with radiographic infarcts in MRI, number of people with clinical brain infarcts, and mean days of MRI performed after each procedure). Any disagreements were resolved by discussion with a third person (K.U.). The data were entered into a standardized pretested electronic form. Article authors were contacted for any information missing from the published article that was deemed relevant or for inquiry and concerns about overlapping cohorts in studies.

For diagnostic cardiac catheterization, we defined high-risk catheterization if diagnostic cardiac catheterization involved crossing of aortic valve with or without aortic stenosis because this carries a higher risk of embolism. We excluded the cardiac catheterization with intervention and also excluded if population is heterogeneous without separating intervention and nonintervention group. Diagnostic cerebral angiograms performed in temporal proximity to acute stroke were excluded because recurrent stroke, not procedural embolism, would add to the risk of infarction. Studies were excluded if postprocedural MRI was performed in a portion of study population with a concern of selection bias.

We utilized the stroke and TIA definitions of each study. When TIA and stroke were not clearly defined in the article, we used the updated definition of stroke and TIA in the American Heart Association/American Stroke Association consensus statement from year 2013.5 Both TIAs and strokes are included in clinical (symptomatic) brain infarcts, knowing that some TIAs may or may not have absent radiographic infarcts because the included studies do not specify this question.

Statistical Analysis, Assessment of Heterogeneity, and Risk of Bias

The main objective is to compare the rates of RBIs and clinical brain infarcts. A ratio between the 2 rates provides such a comparison in a single number. We calculated a symptomatic rate ratio (RR), such that RR=(stroke+TIA) rate/RBI rate. We decided that the RBI would be a denominator given that there were no zeros across the studies for RBI. We used random effects models with the inverse variance method. The RR is expressed with its 95% confidence interval (CI). Differences of RRs across subgroups were tested with a χ2 text. A P<0.05 expressed differences across subgroups. Heterogeneity of RRs was tested with the χ2 test with a P<0.1 indicating presence of heterogeneity. I2 quantified the degree of heterogeneity (high heterogeneity >60%, low heterogeneity <30%, and intermediate heterogeneity between 30% and 60%).

The Newcastle–Ottawa Scale (NOS) was applied to evaluate risk of bias in 3 domains for cohort studies.6 The NOS uses 2 different tools for case–control and cohort studies and consists of 3 parameters of quality: selection, comparability, and exposure/outcome assessment. The NOS assigns a maximum of 4 points for selection, 2 points for comparability, and 3 points for exposure or outcome. NOS scores of ≥7 were considered as having low risk of bias, and NOS scores of 5 to 6 were considered as having moderate risk of bias. Any discrepancies were addressed by a joint re-evaluation of the original article. To assess the risk of bias of randomized controlled trials, we used The cochrane collaboration tools,7each of the 8 items of this tool was judged as high risk, low risk or unclear risk. Studies with high risk of bias in the items related to randomization or blinding were judged as having high risk of bias. Case series studies were not evaluated for risk of bias because their design is of low quality. To check for publication bias, we used Egger regression asymmetry test and trim and fill method.

The RevMan 5.3. software (Copenhagen: The Nordic Cochrane Center, The Cochrane Collaboration, 2014) was used for all analyses.

Results

The literature search yielded 6332 articles (Figure 1), of which 29 studies (2124 patients) met the inclusion criteria (Appendix 2 in the online-only Data Supplement). All studies had a short-term follow-up with MR imaging and clinical neurological examination. Among the 29 studies, there were 17 cohorts in AVR, 4 cohorts in CABG, 10 cohorts in cardiac catheterization, and 8 cohorts in cerebral angiogram. MR DWI scan was performed pre-procedure in 27 studies and post-procedure in all 29 studies. Majority of the studies (24 studies: 83%) utilized 1.5 Tesla (T) MRI; 1 with 3.0 T, 1 with 1.0 T, and 3 studies without MRI information. The reported mean time to postprocedural MRIs was a median of 2 days (interquartile range, 1–6.5).

Figure 1.

Figure 1. Study flowchart for literature search and selection of studies. MRI indicates magnetic resonance imaging; and RBI, radiographic brain infarct.

Risk of Bias Assessment

The NOS quality assessment was conducted on 18 cohort studies. For each meta-analysis of 3 end points, the NOS assessment did not indicate high risk of bias. The median NOS score was 6 (Table II in the online-only Data Supplement). The cochrane collaboration tool showed overall low risk or unclear risk of bias in 5 randomized controlled studies (Table III in the online-only Data Supplement). Egger regression asymmetry test and trim and fill method were performed in AVR and cardiac catheterization subgroups because they have 10 or more studies. For the AVR group, the Egger test for plot asymmetry had a P=0.27 suggesting no asymmetry. The trim and fill method gave similar results to the original analyses suggesting low risk of publication bias. For the cardiac catheterization group, the Egger test for plot asymmetry had a P=0.78 suggesting no asymmetry. The trim and fill method gave similar results to the original analyses suggesting low risk of publication bias.

Meta-Analysis: RR of Clinical-to-Radiographic Infarcts

Table 1 summarizes the rates of clinical infarct and RBI rates for 4 procedures and subtypes. The overall symptomatic RR was 0.10 (95% CI, 0.07–0.13; Figure 2) and did not differ across different procedure categories (P=0.29; I2=0%). In other words, 1 of 10 people with periprocedural RBIs during cardiovascular procedures resulted in clinical brain infarcts. Heterogeneity across subgroups were low (I2=19.4%) and heterogeneity within each procedure type was 0% except I2 of 20% for cerebral angiogram (Table 1).

Table 1. Summary of Meta-Analysis of Data of Clinical and Radiographic Brain Infarcts in Different Procedures

ProceduresNo. of CohortsTotal No. of ProceduresClinical Infarcts, n (%)95% CIRadiographic Infarcts, n (%)95% CI
All AVR1749421 (3.6)2.0–5.2347 (69)58–81
 SAVR51495 (3.4)0.5–6.377 (45)10–80
 TAVR1234516 (3.6)1.7–5.6270 (80)72–89
CABG42045 (2.4)0.3–4.551 (27)6.0–49
All cardiac catheterization108338 (0.6)0.1–1.183 (8.0)4.1–12
 Low-risk cardiac catheterization54430 (0.4)0–1.029 (5.2)1.0–9.4
 High-risk cardiac catheterization53908 (1.6)0.4–2.854 (10.8)3.5–18.1
Cerebral angiogram85934 (0.6)0–1.390 (12.8)6.6–19

AVR indicates aortic valve replacement; CABG, coronary artery bypass grafting; CI, confidence interval; SAVR, surgical (open) aortic valve replacement; and TAVR, transcatheter aortic valve replacement.

Figure 2.

Figure 2. Meta-analysis of rate ratios across 4 subgroups: aortic valve replacement/repair (AVR), coronary artery bypass graft (CABG), cardiac catheterization, and cerebral angiogram. CI indicates confidence interval.

In a meta-analysis of all AVRs, 69.4% (95% CI, 57.6%–81.4%) had infarcts on MRI, whereas 3.6% (95% CI, 2.0%–5.2%) had clinical strokes or TIAs (Table 1) with a RR of 0.08 (95% CI, 0.05–0.12). The reported RR differs from ratio of stroke and TIA percentage divided by infarct rate (ie, 3.6%/69.4%) because the RR is the weighted meta-analysis of ratio in each study. In a subgroup analysis, open AVR had 45.1% (95% CI, 10.0%–80.3%) of RBIs and 3.4% (95% CI, 0.5%–6.3%) of clinical brain infarcts. In TAVR, 80.3% (95% CI, 72.0%–89.0%) had RBIs, whereas 3.7% (95% CI, 1.7%–5.6%) had clinical brain infarcts. Both clinical brain infarct and RBI rates did not differ between SAVR and TAVR (Table 2; P=0.7 and P=0.3).

Table 2. Summary of Comparison Analyses of Subgroups

P Value
Clinical Infarct Rate ComparisonRadiographic Infarct Rate Comparison
SAVR vs TAVR0.70.3
On pump vs off pump CABG0.50.9
AVR vs CABG0.70.01
High-risk vs low-risk cardiac catheterization0.30.5

AVR indicates aortic valve replacement; CABG, coronary artery bypass grafting; SAVR, surgical (open) aortic valve replacement; and TAVR, transcatheter aortic valve replacement.

In CABG, 27.4% (95% CI, 6.0%–48.8%) of patients had RBIs and 2.4% (95% CI, 0.3%–4.5%) had clinical brain infarcts. The RR was 0.11 (95% CI, 0.05–0.26). Comparison analysis between CABG with on-pump versus CABG with off-pump was performed, and the rates of clinical brain infarcts and RBIs were not significantly different between the 2 CABG types with P=0.5 and P=0.9, respectively. This analysis showed a low heterogeneity of RRs within these 2 subgroups (Table 1). Further analysis on CABG versus AVR was conducted to investigate whether certain types of surgery carry higher risk for cerebral infarctions. The clinical brain infarcts did not differ in AVR versus CABG (P=0.7); however, RBIs were significantly higher in AVR when compared with CABG (P=0.01).

In diagnostic cardiac catheterization, 8.0% (95% CI, 4.1%–12.0%) of patients had RBIs and 0.6% (95% CI, 0.1%–1.1%) had clinical brain infarcts. The RR was 0.16 (95% CI, 0.08–0.31) for cardiac catheterization procedures. We further divided this group to high-risk versus low-risk cardiac catheterization. The high-risk and low-risk cardiac catheterization each had 10.8% and 5.2% clinical brain infarcts and 1.6% and 0.4% RBIs, respectively. The clinical brain infarcts and RBIs rates were not significantly different between these 2 subgroups (high risk versus low risk) with P=0.3 and P=0.5, respectively (Table 2).

Diagnostic cerebral angiogram had 12.8% (95% CI, 6.6%–19.0%) RBIs and 0.6% (95% CI, 0%–1.3%) clinical brain infarcts. The RR was 0.10 (95% CI, 0.04–0.27) for diagnostic cerebral angiogram.

In addition, exploratory subgroup analysis was performed to see whether the RRs in 2 cardiac surgeries (AVR and CABG) differ from 2 diagnostic procedures (cardiac catheterization and diagnostic cerebral angiogram), and there was no difference in the RRs of AVR plus CABG versus cerebral angiogram plus cardiac catheterization subgroups (P=0.16; Figure I in the online-only Data Supplement). Also, the symptomatic infarct ratios were compared by the year of publication (group 1=1996–2009 versus group 2=2010–2015) by choosing the median year (2009) of 49 cohorts to divide into 2 subgroups. There was no difference in RRs between group 1 and group 2: 0.13 (95% CI, 0.08–0.23) versus 0.09 (95% CI, 0.06–0.12), P=0.21 (Figure II in the online-only Data Supplement).

Discussion

Our primary aim was to assess the ratio between clinical infarcts and RBIs to see whether there is a consistent infarct ratio across different cardiac surgeries and diagnostic procedures. We calculated a symptomatic RR, defined as strokes+TIAs rate divided by RBIs rate, for the analysis. Our analysis was consistent with the hypothesis in that there is a consistent infarct ratio across the procedures, a symptomatic RR of 0.10 with low heterogeneity. In other words, in 29 studies with a total of 2124 patients, in ≈1 of 10 people with radiographic infarcts resulted in stroke or TIA symptoms. This ratio did not differ among different types of cardiac and diagnostic procedures (AVR, CABG, cardiac catheterization, and cerebral angiogram) or different times of publications (2010–2015 versus 1996–2009), which do not indicate that the ratio is heterogeneous, and we can expect a uniform percentage of procedural cerebral embolism to be symptomatic. Although we have known that clinically silent brain infarcts are common after invasive vascular procedures, clinical utility of this information has been limited. We propose that this ratio can be utilized in studies of procedural safety using MRIs and as a predictor of risk for periprocedural ischemic embolic events.

Interestingly, this constant symptomatic ratio of periprocedural infarcts is similar to the symptomatic ratios of chronic silent brain infarction reported in population-based studies. In a cohort of 1433 with 2 MRI scans with 5-year follow-up, 29 (11%) of 254 people with MRI infarcts had TIA or stroke in clinical history between the 2 MRI scans.8 Another cohort of 1077 people was followed for a mean of 3.4 years with 2 MRIs and 93 people had new infarcts on MRI, of which 12 people (13%) had strokes or TIAs.9 The similarity of the symptomatic brain infarct ratios between the silent brain infarction over time in the community and our study of acute periprocedural infarctions suggest the same underlying robustness or complexity of the brain that limits obvious symptoms. But subtle cognitive decline can be detected among people with acute periprocedural DWI lesions than those without, and chronic silent brain infarctions similarly are associated with increased risk of dementia.10,11 Although we found symptomatic ratio of 0.1, we recognize that routine clinical examination would have picked up only classic stroke symptoms, and the quality of these examinations cannot be ascertained in this systematic review.

Isolated AVR and isolated CABG carry a relatively low incidence of perioperative stroke with 1.5% to 4% and 1% to 3% in AVR and CABG, respectively.1218 To compare periprocedural cerebral embolic complication risks to reduce them, event rates of <5% would require a large cohort of subjects. Studying a biomarker that has 10× the event rate would reduce the needed sample size. MRI is being used as a surrogate marker in cerebral protection device studies.19

Compared with open AVR, it was suggested that TAVR carries a higher risk of stroke (1.5%–6%) in retrospective studies,17 and 1 randomized controlled trial showed an increased risk of strokes at day 30 days in TAVR compared with SAVR (6.7% versus 1.7%) in high-risk population.20 More recent study of intermediate risk population with newer device showed a similar risk of 30-day strokes in TAVR and SAVR.21 Our analysis of nonrandomized samples showed a similar rate of strokes and radiographic infarct rates between TAVR and SAVR. The reason for the lower incidence of stroke rates in our analysis might be because of early neurological assessment and MRI (mean time to postprocedural MRI scan within a median of 4.5 days), whereas the large clinical trials report 30-day complication rates.

In CABG, similar to AVR, low rate of clinical infarct (2.4%) was observed, and RBIs were present in 27.4% (95% CI, 6.0%–48.8%) of people with RR of 0.11, demonstrating again ≈1 of 10 people with periprocedural RBIs resulted in neurologically symptomatic brain infarcts. Furthermore, the comparison analysis between CABG with on-pump versus off-pump showed comparable rates of clinical infarcts and RBIs, consistent with prior randomized controlled trial.22

Diagnostic procedures carry a low risk of stroke (0.0%–1.0%) in cerebral angiogram23,24 and cardiac catheterization.2527 The ischemic embolic events are thought to be because of dislodged embolic debris from atherosclerotic plaques in aortic arch or major extra- and intracranial vessels. Therefore, it has been suggested that retrograde cardiac catheterization across an aortic valve may increase the risk of embolism, but our study did not find significant increase in risk.28

A large number of studies and patients included for the analyses is the strength of our study. However, our study has several obvious limitations. Across the large number of studies, clearly the procedural details differ. But the goal of the study was to see the robustness of the ratio, despite variations in infarct and stroke rates. When the comparisons are infarct rates among different procedures, our comparisons across studies are limited by differences in procedural details, operators, and changes in procedural methodology that may occur over time. We are unable to account for differences in MRI methods across studies, including magnetic field strength, timing post-procedure, and interpretation that might affect sensitivity of DWI. Also, we have no information about the relationship between radiographic infarct numbers, size, and location in each patient to their symptomatic status because the images were not available.

Conclusions

One of 10 people with periprocedural RBIs during cardiac surgeries and invasive vascular diagnostic procedures resulted in strokes or TIAs, which may serve as a potential surrogate marker of procedural proficiency and perhaps as a predictor of risk for periprocedural strokes.

Acknowledgments

S.-M. Cho, Dr Deshpande, and Dr Uchino contributed to the study concept and design. S.-M. Cho prepared the first draft of the article. Drs Deshpande, Hernandez, and Uchino contributed in drafting the article. S.-M. Cho and Dr Uchino finalized the article. All authors contributed to data acquisition and analysis and approved the final version of the article.

Footnotes

The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.117.017541/-/DC1.

Reprint requests to Ken Uchino, MD, FAHA, FANA, Cerebrovascular Center, Neurological Institute, Cleveland Clinic, 9500 Euclid Ave S80, Cleveland, OH 44195. E-mail

References

  • 1. Rosso C, Drier A, Lacroix D, Mutlu G, Pires C, Lehéricy S, et al.. Diffusion-weighted MRI in acute stroke within the first 6 hours: 1.5 or 3.0 Tesla?Neurology. 2010; 74:1946–1953. doi: 10.1212/WNL.0b013e3181e396d1.CrossrefMedlineGoogle Scholar
  • 2. Brazzelli M, Sandercock PA, Chappell FM, Celani MG, Righetti E, Arestis N, et al.. Magnetic resonance imaging versus computed tomography for detection of acute vascular lesions in patients presenting with stroke symptoms.Cochrane Database Syst Rev. 2009; 88:S23–S42. doi: 10.1002/14651858.CD007424.pub2.Google Scholar
  • 3. Bendszus M, Stoll G. Silent cerebral ischaemia: hidden fingerprints of invasive medical procedures.Lancet Neurol. 2006; 5:364–372. doi: 10.1016/S1474-4422(06)70412-4.CrossrefMedlineGoogle Scholar
  • 4. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement.PLoS Med. 2009; 6:e1000097. doi: 10.1371/journal.pmed.1000097.CrossrefMedlineGoogle Scholar
  • 5. Sacco RL, Kasner SE, Broderick JP, Caplan LR, Connors JJ, Culebras A, et al.; American Heart Association Stroke Council, Council on Cardiovascular Surgery and Anesthesia; Council on Cardiovascular Radiology and Intervention; Council on Cardiovascular and Stroke Nursing; Council on Epidemiology and Prevention; Council on Peripheral Vascular Disease; Council on Nutrition, Physical Activity and Metabolism. An updated definition of stroke for the 21st century: a statement for healthcare professionals from the American Heart Association/American Stroke Association.Stroke. 2013; 44:2064–2089. doi: 10.1161/STR.0b013e318296aeca.LinkGoogle Scholar
  • 6. Deeks JJ, Dinnes J, D’Amico R, Sowden AJ, Sakarovitch C, Song F, et al.; International Stroke Trial Collaborative Group; European Carotid Surgery Trial Collaborative Group. Evaluating non-randomised intervention studies.Health Technol Assess. 2003; 7:iii–x, 1–173.CrossrefMedlineGoogle Scholar
  • 7. Higgins JP, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al.; Cochrane Bias Methods Group; Cochrane Statistical Methods Group. The cochrane collaboration’s tool for assessing risk of bias in randomised trials.BMJ. 2011; 343:d5928.CrossrefMedlineGoogle Scholar
  • 8. Longstreth WT, Dulberg C, Manolio TA, Lewis MR, Beauchamp NJ, O’Leary D, et al.. Incidence, manifestations, and predictors of brain infarcts defined by serial cranial magnetic resonance imaging in the elderly: the Cardiovascular Health Study.Stroke. 2002; 33:2376–2382.LinkGoogle Scholar
  • 9. Vermeer SE, Den Heijer T, Koudstaal PJ, Oudkerk M, Hofman A, Breteler MM; Rotterdam Scan Study. Incidence and risk factors of silent brain infarcts in the population-based Rotterdam scan study.Stroke. 2003; 34:392–396.LinkGoogle Scholar
  • 10. Fanning JP, Wesley AJ, Walters DL, Eeles EM, Barnett AG, Platts DG, et al.. Neurological injury in intermediate-risk transcatheter aortic valve implantation.J Am Heart Assoc. 2016; 5:e004203.LinkGoogle Scholar
  • 11. Vermeer SE, Prins ND, den Heijer T, Hofman A, Koudstaal PJ, Breteler MM. Silent brain infarcts and the risk of dementia and cognitive decline.N Engl J Med. 2003; 348:1215–1222. doi: 10.1056/NEJMoa022066.CrossrefMedlineGoogle Scholar
  • 12. Roach GW, Kanchuger M, Mangano CM, Newman M, Nussmeier N, Wolman R, et al.. Adverse cerebral outcomes after coronary bypass surgery. Multicenter study of perioperative ischemia research group and the ischemia research and education foundation investigators.N Engl J Med. 1996; 335:1857–1863. doi: 10.1056/NEJM199612193352501.CrossrefMedlineGoogle Scholar
  • 13. Tarakji KG, Sabik JF, Bhudia SK, Batizy LH, Blackstone EH. Temporal onset, risk factors, and outcomes associated with stroke after coronary artery bypass grafting.JAMA. 2011; 305:381–390. doi: 10.1001/jama.2011.37.CrossrefMedlineGoogle Scholar
  • 14. Nah HW, Lee JW, Chung CH, Choo SJ, Kwon SU, Kim JS, et al.. New brain infarcts on magnetic resonance imaging after coronary artery bypass graft surgery: lesion patterns, mechanism, and predictors.Ann Neurol. 2014; 76:347–355. doi: 10.1002/ana.24238.CrossrefMedlineGoogle Scholar
  • 15. Vasques F, Messori A, Lucenteforte E, Biancari F. Immediate and late outcome of patients aged 80 years and older undergoing isolated aortic valve replacement: a systematic review and meta-analysis of 48 studies.Am Heart J. 2012; 163:477–485. doi: 10.1016/j.ahj.2011.12.005.CrossrefMedlineGoogle Scholar
  • 16. O’Brien SM, Shahian DM, Filardo G, Ferraris VA, Haan CK, Rich JB, et al.; Society of Thoracic Surgeons Quality Measurement Task Force. The society of thoracic surgeons 2008 cardiac surgery risk models: part 2–isolated valve surgery.Ann Thorac Surg. 2009; 88(suppl 1):S23–S42. doi: 10.1016/j.athoracsur.2009.05.056.CrossrefMedlineGoogle Scholar
  • 17. Daneault B, Kirtane AJ, Kodali SK, Williams MR, Genereux P, Reiss GR, et al.. Stroke associated with surgical and transcatheter treatment of aortic stenosis: a comprehensive review.J Am Coll Cardiol. 2011; 58:2143–2150. doi: 10.1016/j.jacc.2011.08.024.CrossrefMedlineGoogle Scholar
  • 18. Thourani VH, Ailawadi G, Szeto WY, Dewey TM, Guyton RA, Mack MJ, et al.. Outcomes of surgical aortic valve replacement in high-risk patients: a multiinstitutional study.Ann Thorac Surg. 2011; 91:49–55; discussion 55–56. doi: 10.1016/j.athoracsur.2010.09.040.CrossrefMedlineGoogle Scholar
  • 19. Kapadia SR, Kodali S, Makkar R, Mehran R, Lazar RM, Zivadinov R, et al.. Cerebral embolic protection during transcatheter aortic valve replacement.J Am Coll Cardiol. 2017; 69:367–377. doi: 10.1016/j.jacc.2016.10.023.CrossrefMedlineGoogle Scholar
  • 20. Leon MB, Smith CR, Mack M, Miller DC, Moses JW, Svensson LG, et al.; PARTNER Trial Investigators. Transcatheter aortic-valve implantation for aortic stenosis in patients who cannot undergo surgery.N Engl J Med. 2010; 363:1597–1607. doi: 10.1056/NEJMoa1008232.CrossrefMedlineGoogle Scholar
  • 21. Leon MB, Smith CR, Mack MJ, Makkar RR, Svensson LG, Kodali SK, et al.; PARTNER 2 Investigators. Transcatheter or surgical aortic-valve replacement in intermediate-risk patients.N Engl J Med. 2016; 374:1609–1620. doi: 10.1056/NEJMoa1514616.CrossrefMedlineGoogle Scholar
  • 22. Shroyer AL, Grover FL, Hattler B, Collins JF, McDonald GO, Kozora E, et al.; Veterans Affairs Randomized On/Off Bypass (ROOBY) Study Group. On-pump versus off-pump coronary-artery bypass surgery.N Engl J Med. 2009; 361:1827–1837. doi: 10.1056/NEJMoa0902905.CrossrefMedlineGoogle Scholar
  • 23. Bendszus M, Koltzenburg M, Bartsch AJ, Goldbrunner R, Günthner-Lengsfeld T, Weilbach FX, et al.. Heparin and air filters reduce embolic events caused by intra-arterial cerebral angiography: a prospective, randomized trial.Circulation. 2004; 110:2210–2215. doi: 10.1161/01.CIR.0000144301.82391.85.LinkGoogle Scholar
  • 24. Brockmann C, Hoefer T, Diepers M, Neumaier-Probst E, Noelte I, Brockmann MA, et al.. Abciximab does not prevent ischemic lesions related to cerebral angiography: a randomized placebo-controlled trial.Cerebrovasc Dis. 2011; 31:353–357. doi: 10.1159/000323219.CrossrefMedlineGoogle Scholar
  • 25. Dukkipati S, O’Neill WW, Harjai KJ, Sanders WP, Deo D, Boura JA, et al.. Characteristics of cerebrovascular accidents after percutaneous coronary interventions.J Am Coll Cardiol. 2004; 43:1161–1167. doi: 10.1016/j.jacc.2003.11.033.CrossrefMedlineGoogle Scholar
  • 26. Werner N, Bauer T, Hochadel M, Zahn R, Weidinger F, Marco J, et al.. Incidence and clinical impact of stroke complicating percutaneous coronary intervention: results of the Euro Heart Survey percutaneous coronary interventions registry.Circ Cardiovasc Interv. 2013; 6:362–369. doi: 10.1161/CIRCINTERVENTIONS.112.000170.LinkGoogle Scholar
  • 27. Kwok CS, Kontopantelis E, Myint PK, Zaman A, Berry C, Keavney B, et al.; British Cardiovascular Intervention Society; National Institute for Cardiovascular Outcomes Research. Stroke following percutaneous coronary intervention: type-specific incidence, outcomes and determinants seen by the British cardiovascular intervention society 2007-12.Eur Heart J. 2015; 36:1618–1628. doi: 10.1093/eurheartj/ehv113.CrossrefMedlineGoogle Scholar
  • 28. Omran H, Schmidt H, Hackenbroch M, Illien S, Bernhardt P, von der Recke G, et al.. Silent and apparent cerebral embolism after retrograde catheterisation of the aortic valve in valvular stenosis: a prospective, randomised study.Lancet. 2003; 361:1241–1246. doi: 10.1016/S0140-6736(03)12978-9.CrossrefMedlineGoogle Scholar