Predicting Cardioembolic Stroke with the B-Type Natriuretic Peptide Test: A Systematic Review and Meta-analysis Hong-ling Yang, PhD,* Yong-Ping Lin, MD, PhD,†‡ Yan Long, PhD,* Qing-ling Ma, MD,* and Cheng Zhou, MD*

Background: We performed a systematic review and meta-analysis to evaluate the value of B-type natriuretic peptide (BNP) in differentiating cardioembolic (CE) stroke from other subtypes of ischemic stroke. Methods: We searched the EMBASE, MEDLINE, and Cochrane databases and reference lists of relevant articles published in April 2013. We selected original studies reporting the performance of BNP or N-terminal probrain natriuretic peptide (NT-proBNP) in diagnosing CE stroke and summarized test performance characteristics using forest plots, hierarchical summary receiver operating characteristic curves, and bivariate random-effect models. Results: Data from 2958 patients with ischemic stroke were retrieved from 16 studies. Of these, 1024 (34.6%) patients had a final diagnosis of CE stroke. Overall, the mean diagnostic odds ratio (DOR) of BNP for CE stroke was 15.8 (95% confidence interval [CI]: 9.92-25.20). Even after adjustment for multiple clinical predictors, serum natriuretic peptide levels showed a strong association with CE stroke (pooled adjusted DOR, 12.7; 95% CI: 7.32-22.0). The sensitivity and specificity of BNP for CE stroke were .78 (95% CI: .71-.87) and .83 (95% CI: .77-.87), respectively. A single BNP-negative result may be sufficient to exclude a diagnosis of CE stroke in low-prevalence (,20%) settings. Subgroup analysis showed that NT-proBNP had a slightly higher specificity (.87; 95% CI: .77-.93) and better capability for exclusion diagnosis. There was a lack of homogeneity in the timing of measurement and BNP assay method. Conclusions: BNP has reasonable accuracy in the diagnosis of CE stroke and may be a useful marker for the early detection in patients who may benefit from preventive anticoagulation therapy. Key Words: Embolism— cardioemboli—B-type natriuretic peptide—stroke. Ó 2014 by National Stroke Association

From the *Department of Clinical Laboratory, Guangzhou Women and Children’s Medical Center, Guangzhou Medical University; †Department of Laboratory Medicine, The First Affiliated Hospital of Guangzhou Medical University, Guangdong; and ‡Department of Translational Research Center, The First Affiliated Hospital of Guangzhou Medical University, Guangdong, China. Received February 11, 2014; accepted February 21, 2014. The authors declare no conflicts of interest. Address correspondence to Yong-Ping Lin, MD, PhD, No. 151, Yanjiang Xi Road, Yuexiu District, Guangzhou City, Guangdong 510120, China. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.02.014

Stroke is the leading cause of disability and the third leading cause of death in developed countries.1,2 Cardioembolic (CE) stroke accounts for 20% of all cerebral infarctions, and it is the most common subtype among all strokes with an identifiable cause.3-5 Cardiac emboli arising from the cardiac chambers are often large; therefore, CE infarction is generally the most severe ischemic stroke subtype, with a higher rate of disability or death. Because CE recurrence is largely preventable by oral anticoagulation therapy, early confirmation of a diagnosis of CE infarction among patients with acute ischemic stroke is extremely important.2,4 There is no gold standard criterion for the diagnosis of CE stroke. The Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria classify patients with ischemic

Journal of Stroke and Cerebrovascular Diseases, Vol. -, No. - (---), 2014: pp 1-8

1

H.-L. YANG ET AL.

2

stroke into 5 core etiologic groupings on the basis of risk factor profiles, clinical features, and diagnostic test findings.6 The 5 subtypes of stroke include large-artery atherosclerosis, cardioembolism, small artery occlusion, stroke of other determined cause, and stroke of undetermined cause. Diagnosis of CE stroke is based on the presence of a potential major cardiac source of embolism after the exclusion of severe arterial disease, such as arteritis, arterial dissection, or venous occlusion. Left atrial appendage in patients with atrial fibrillation (AF) is a major source of cardioembolism.7 Other relatively infrequent sources include thrombus formation from an abnormal valvular surface or paradoxical venous embolism from a right to left shunt. Persistent AF, recent myocardial infarction, acute heart failure, or splinter hemorrhages provide initial clinical clues for the diagnosis of cardiac emboli.4 Transthoracic or transesophageal echocardiography can confirm the presence of cardiac emboli or disclose structural changes in the heart that can promote cardiac emboli formation.8 However, even after extensive investigation, the cause of stroke remains undetermined in approximately 30%40% patients.8,9 Paroxysmal AF has been suggested to account for a significant proportion in patients with undetermined stroke etiology.10 B-type natriuretic peptide (BNP) or its precursor, N-terminal probrain natriuretic peptide (NT-proBNP), is a marker currently used for diagnostic and prognostic purposes in patients with heart failure.11 Many studies have shown that the left atrium is the main source of BNP in patients with AF, and it is believed that this marker can help in the early detection of new-onset AF. Recently, several studies evaluated the value of BNP in differentiating CE stroke from other subtypes of ischemic stroke.12-24 However, the results of these studies have been inconsistent. We, therefore, undertook a systematic review and conducted a meta-analysis to evaluate the performance of BNP/NT-proBNP assays in differentiating CE stroke from other subtypes of stroke.

Materials and Methods This systematic review and meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines and the methods and procedures of the Cochrane Collaboration for reporting systematic reviews.25,26 We performed a comprehensive search of the MEDLINE, EMBASE, and Cochrane databases; BIOSIS; and Web of Science for pertinent studies published since inception in April 2013. Natriuretic peptide (NP) has not yet been listed as a Medical Subject Headings term; therefore, for our initial search, we used ‘‘natriuretic peptide,’’ ‘‘B-type natriuretic peptide,’’ ‘‘N-terminal pro–Btype natriuretic peptide,’’ ‘‘BNP,’’ or ‘‘NT-proBNP’’ to cross with ‘‘stroke.’’ The same strategy was used with

EMTREE tools to search the EMBASE database without setting any time or language restrictions. We also identified additional references by cross-checking bibliographies of retrieved full-text articles and contacted authors for additional data. Initial eligibility was independently determined by 2 reviewers. Disagreements between reviewers were resolved by additional reviewers in a consensus meeting. We included original clinical studies that met all the following criteria: (1) older than 20 years and a history of acute ischemic stroke, (2) evaluation of BNP or NTproBNP to detect CE stroke, and (3) availability of sufficient data to construct a 2 3 2 contingency table. We excluded editorials, perspectives, opinion pieces, manufacturer reports, and studies in other specimens. Studies that compared healthy controls were not eligible for inclusion because they tend to overestimate sensitivity and specificity.

Data Extraction We used a structured data abstraction form with variables, such as overall study characteristics (including the first author, country, language, and date of publication), patient characteristics (including age range and percentage of male patients), quantitative data required for the construction of a 2 3 2 table (including number of participants, sensitivity, specificity, and case number), information about the NP test (including cutoff levels, timing of measurement, and type of NP measured), study settings, and outcomes.

Quality Assessment of Included Studies We assessed methodological quality using the quality assessment for studies of diagnostic accuracy tool.27 The spectrum of patients included in a study was considered to be a representative of the target population if they had a clinical diagnosis of stroke. The reference standard used was the TOAST stroke classification. Partial and differential verification biases were considered if all the included patients were not assessed with the same reference standard. Incorporation bias was considered if the diagnosis of the subtype of stroke was made with the knowledge of serum BNP or NT-proBNP levels.

Data Analysis We used a bivariate meta-analysis model to calculate the mean sensitivity, specificity, positive likelihood ratio (LR), negative LR, and odds ratio for diagnosing CE stroke.28 This model overcomes the limitations of the traditional model by taking into account the negative correlation between sensitivity and specificity and by incorporating between-study heterogeneity in both parameters. We constructed a hierarchical summary receiver operating characteristic (ROC) curve plotting

PREDICTING CARDIOEMBOLIC STROKE WITH THE BNP

Patient characteristics, study settings, proportion of patients with a final diagnosis of CE stroke, tested biomarkers, cutoff values, and correspondent sensitivity

ED Inpatient Inpatient Inpatient Inpatient Inpatient NA ED Inpatient Stroke unit Inpatient ED ED ED 68, 72 76, 77 97, 98 87, 83 81, 81 74, 88 38, 98 71, 74 85, 78 63, 75 81, 63 78, 90 93, 75 75, 89 76 77 65 360 140 140 100 265.5 90 64 155 265.5 342 66.5 BNP BNP BNP NT-proBNP BNP BNP NT-proBNP NT-proBNP BNP BNP BNP NT-proBNP NT-proBNP BNP Prospective cohort Prospective cohort Prospective cohort Prospective cohort Prospective cohort Prospective cohort Prospective cohort Prospective cohort Prospective cohort Prospective cohort Prospective cohort Prospective cohort Prospective cohort Prospective cohort 72 6 12 69.6 6 10.1 70.8 6 11.5 69.2 71.4 72.9 72.1 6 10.4 58.6 69 6 14 71.32 6 13 71.1 6 11.5 64.5 6 12.3 67.56 6 12.6 70.0 6 10.1

Biomarkers tested Design

Abbreviations: BNP, B-type natriuretic peptide; ED, emergency department; NT-proBNP, N-terminal probrain natriuretic peptide.

Study Characteristics

.37 (707) .47 (131) .47 (76) .38 (262) .41 (200) .35 (221) .30 (100) .42 (66) .31 (223) .17 (294) .35 (310) .29 (101) .46 (125) .25 (142)

Our search yielded 1774 unique results. After screening titles and abstracts, 38 articles were selected for full-text review. We did not find additional eligible studies after reviewing the bibliographies of the included articles. After screening the full text of these 38 articles, 14 were finally included for analysis. A flow chart of the inclusion process is shown in Figure 1.

Montaner et al,38 Spain Yukiiri et al,23 Japan Naya et al,16 Japan Rodriguez-Yanez et al,18 Spain Shibazaki et al,21 Japan Sakai et al,19 Japan Okada et al,17 Japan Fonseca et al,13 Portugal Tamura et al,22 Japan Santamarina et al,20 Spain Biteker et al,12 Turkey Fonseca et al,14 Portugal Hajsadeghi et al,15 Iran Zhixin et al,24 China

Results

Mean age

sensitivity vs specificity and calculated the area under the curve.29 We plotted the confidence regions for the summary points and the prediction region of 95% of future studies. To deal with values of zero in the 2 3 2 contingency tables, we performed continuity correction, thereby decreasing small study bias. We evaluated the degree of interstudy statistical heterogeneity using the I2 test.30 Subgroup analyses were performed on studies using the NT-proBNP or BNP assays. We tested the publication bias using Egger test, which uses regression methods to test asymmetry of funnel plots.31 Skewed and asymmetrical funnel plots indicate the presence of publication bias. For all the previously mentioned analyses, we used STATA statistical software, version 11.0 (Stata Corp, College Station, TX). All statistical tests were 2 tailed, and statistical significance was defined as a P value less than .05.

Prevalence (n)

Flow chart of study identification and inclusion.

Reference, year

Figure 1.

Table 1. Characteristics of the patients in the included studies

Cutoff (pg/mL)

Sensitivity (%), specificity (%)

Setting

3

H.-L. YANG ET AL.

4

and specificity are summarized in Table 1. The 14 included studies totaled 2958 patients, with 1024 (34.6%) having a final diagnosis of CE stroke. Optimal cutoff values ranged from 64 to 155 pg/mL for BNP and 100 to 366 pg/mL for NT-proBNP. All studies used an appropriate prospective design. The timing of NP testing ranged from admission to 72 hours. Five studies evaluated BNP tests and 9 evaluated NT-proBNP tests. Five studies were performed in the emergency department, and the rest were performed in inpatient units.

Quality of Included Studies All studies used independent reference tests to verify CE stroke, with little possibility of differential outcome ascertainment. None of the included studies provided an explanation for uninterpretable test results. The reason for participants’ withdrawal from the study and whether clinicians were blinded to the BNP/NT-proBNP test results while ascertaining outcome was not reported in most studies. However, incorporation bias is less likely because the diagnosis of cardioembolism largely depends on objective findings in transthoracic or transesophageal echocardiography. Figure 2 provides an overall picture of the methodological quality of studies as evaluated by the quality assessment for studies of diagnostic accuracy tool.

Diagnostic Accuracy The overall sensitivity and specificity of NP testing for diagnosing CE stroke as reported in the 14 included studies are shown in Table 2. A pooled analysis resulted

Figure 2.

in a sensitivity of 78% (95% confidence interval [CI]: 71%-87%) and a specificity of 83% (95% CI: 77%-87%). In the 5 studies that evaluated the diagnostic performance of NT-proBNP, the sensitivity remained unchanged at 78% (95% CI: 58%-90%), whereas the specificity increased significantly to 87% (95% CI: 77%93%). The pooled positive LR increased from 4.51 (95% CI: 3.31-6.14) to 5.92 (95% CI: 3.78-9.28), whereas the pooled negative LR remained similar (.26, 95% CI: .19.36), making it a better rule-in tool. In the 9 studies evaluating the diagnostic performance of BNP, the sensitivity remained similar at 79% (95% CI: 71%-85%), whereas the specificity decreased to 81% (95% CI: 73%-87%). The pooled positive LR decreased to 4.18 (95% CI: 1.15-4.04), whereas the pooled negative LR remained unchanged (.25, 95% CI: .19-.37). The Galbraith plot did not find any study as a significant outlier. The hierarchical summary ROC curve showed a higher discrimination for NT-proBNP than for BNP. The area under the ROC curve was .90 (95% CI: .87-.92) for NTproBNP, which was slightly larger than .87 (95% CI: .83.89) for BNP (Fig 3). Forest plots also showed a higher pooled diagnostic odds ratio (DOR) and lower heterogeneity for NT-proBNP than for BNP (Fig 4). Eight studies performed multivariate analysis to investigate the independent predictive value of BNP/NT-proBNP. After adjustment for multiple clinical predictors, such as age, gender, AF, and renal impairment in multivariate analysis, serum NP levels still showed a strong association with CE stroke (adjusted odds ratio: 12.7, 95% CI: 7.3222.0). There was no evidence of publication bias (Table 2).

Quality of included studies.

PREDICTING CARDIOEMBOLIC STROKE WITH THE BNP

5

.001 .688 .002 79.5 (66.3-87.5) 40.5 (.00-78.0) 80.6 (64.1-89.5) 15.8 (9.92-25.2) 24.6 (15.5-39.1) 13.0 (7.64-22.1) Abbreviations: AUROC, area under receiver operating characteristic; CI, confidence interval; OR, odds ratio.

.88 (.84-.90) .90 (.87-.92) .87 (.83-.89) .26 (.19-.36) .25 (.13-.49) .26 (.19-.37) 4.51 (3.31-6.14) 5.92 (3.78-9.28) 4.18 (2.77-6.31) .83 (.77-.87) .87 (.77-.93) .81 (.73-.87) 14 5 9 Overall12-24,38 NT-proBNP13-15,17,18 BNP10,12,16,19-24

.78 (.71-.87) .78 (.58-.90) .79 (.72-.84)

Diagnostic OR (95% CI) AUROC (95% CI) Likelihood ratio (2) Likelihood ratio (1) Specificity (95% CI) Sensitivity (95% CI) Number of studies Variables

Table 2. Summary of subgroup analysis of the included studies by different study characteristics

I2 (95% CI)

Publication bias (Egger test P)

Discussion Each year, 5 million individuals die as a consequence of stroke, and CE stroke is associated with the highest mortality and highest recurrence rate, with 15% patients experiencing recurrence within 5 years.2,32,33 Because cardioembolism is the most amenable to anticoagulation prevention, early identification of cardioembolism among patients with ischemic stroke is important.4 In this meta-analysis comprising 2958 patients with acute ischemic stroke, we showed that elevated plasma BNP or NT-proBNP levels obtained within 72 hours of stroke onset can help in the early differentiation of CE stroke from other stroke subtypes. NT-proBNP tests have a slightly better specificity and overall accuracy compared with BNP tests. Embolism of cardiac origin accounts for approximately 14%-30% of ischemic stroke cases, and 50% of CE strokes are a result of left atrial thrombus formation because of AF.4,5 Other major sources of embolism include left ventricular thrombus formation in patients with myocardial infarction, dilated cardiomyopathy, or valvular heart disease.8 BNP/NT-proBNP has been shown to be elevated in these conditions, explaining its independent diagnostic value after adjusting for AF in multivariate analysis.34 In addition to the heart, ischemic brain tissue may also be another source of BNP/NTproBNP release, as indicated by the strong correlation between elevated plasma BNP/NT-proBNP levels and poor outcome in stroke patients.35,36 Traditionally, the diagnosis of CE stroke is based on compatible structural heart disease documented by transthoracic or transesophageal echocardiography or atrial or ventricular dysrhythmia detected by Holter monitoring. Echocardiography has a restricted field for thrombus visualization and is heavily dependent on the experience and skill of the operator, limiting its clinical use in the emergency department, where most stroke patients present at symptom onset.8 The BNP/NT-proBNP test can be performed quickly, and it enables the identification of CE stroke early in the emergency department. Early differentiation of stroke subtypes is not useful for secondary prevention. Recent studies have shown that different subtypes of stroke may be associated with different outcomes of thrombolysis treatment. In a study of 72 stroke patients undergoing thrombolytic therapy, CE stroke was associated with more frequent, more complete, and faster early recanalization during tissue plasminogen activator infusion compared with large-vessel diseases and diseases of undetermined origin.37 A better response to thrombolysis treatment may ultimately lead to favorable outcomes. In addition, recent studies have shown that elevated NT-proBNP levels may identify a subgroup of ischemic stroke patients without known AF, and these patients may benefit more from anticoagulants than from antiplatelet agents. These findings suggest that future

H.-L. YANG ET AL.

6

Figure 3. ROC curve for BNP (A) and NT-proBNP (B).

studies investigating different BNP/NT-proBNP–guided thrombolytic or secondary prevention therapies would be beneficial. Other than BNP, several biomarkers have been observed to be elevated in patients with CE stroke. Ddimer, fibrin monomer complex, soluble fibrin, and caspase-3 have all been recently reported to predict CE stroke.20,38 Among these markers, D-dimer was the most commonly studied marker. In most series, D-dimer showed a high sensitivity and low specificity in diagnosing CE stroke; this can be used to complement the low sensitivity of NT-proBNP. Because of the lack of large prospective head-to-head comparison studies, the comparative accuracy between D-dimer and BNP/NTproBNP in terms of CE stroke diagnosis remains undetermined. Furthermore, it would be interesting to determine if a 2-marker strategy (D-dimer for screening and BNP/ NT-proBNP for confirmation) would further enhance the accuracy of biomarker tests for a diagnosis of CE stroke. In addition to a multimarker strategy, clinical predictors may also help in the early diagnosis of CE stroke.

Figure 4.

A model combining clinical variables, such as age, National Institutes of Health Stroke Scale score, AF, embolic cardiopathy, ischemic cardiopathy, and BNP and D-dimer levels, raised the specificity of predicting cardioembolism to 91.3%. However, the sensitivity remained low at 66.5%.39 The main strength of this meta-analysis is that it provides clinically relevant measures, such as sensitivity, specificity, and LRs, other than DOR. However, several limitations need to be addressed. First, the classification of stroke subtype according to the TOAST criteria does not have a gold standard, for example, pathologic examination, for confirmation of the exact etiology. Second, a moderate degree of heterogeneity was noted. Potential sources of heterogeneity included study settings, cutoff levels, and the timing of the BNP/NT-proBNP tests. High heterogeneity among the included studies may limit the generalizability of the studies. Different threshold values for BNP and NT-proBNP assays were used to represent an abnormal value. Because this meta-analysis was not based on individual patient data, it was not

Diagnostic odds ratio (DOR) for BNP (A) and NT-proBNP (B).

PREDICTING CARDIOEMBOLIC STROKE WITH THE BNP

possible to determine an optimal cutoff value across studies. We listed the cutoff values used in individual studies as reference information.

Conclusions This systematic review shows that BNP/NT-proBNP has a modest sensitivity and high specificity for differentiating CE stroke from other stroke subtypes. BNP has a similar sensitivity but lower specificity compared with NT-proBNP. Future studies should focus on increasing the sensitivity for the detection of CE stroke.

References 1. Murray CJ, Lopez AD. Mortality by cause for eight regions of the world: Global Burden of Disease Study. Lancet 1997;349:1269-1276. 2. Strong K, Mathers C, Bonita R. Preventing stroke: saving lives around the world. Lancet Neurol 2007;6:182-187. 3. White H, Boden-Albala B, Wang C, et al. Ischemic stroke subtype incidence among whites, blacks, and Hispanics: the Northern Manhattan Study. Circulation 2005; 111:1327-1331. 4. Arboix A, Alio J. Acute cardioembolic stroke: an update. Expert Rev Cardiovasc Ther 2011;9:367-379. 5. Thom T, Haase N, Rosamond W, et al. Heart disease and stroke statistics—2006 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation 2006;113:e85-151. 6. Goldstein LB, Jones MR, Matchar DB, et al. Improving the reliability of stroke subgroup classification using the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria. Stroke 2001;32:1091-1098. 7. Tagawa M, Takeuchi S, Chinushi M, et al. Evaluating patients with acute ischemic stroke with special reference to newly developed atrial fibrillation in cerebral embolism. Pacing Clin Electrophysiol 2007;30:1121-1128. 8. Sipola P, Hedman M, Onatsu J, et al. Computed tomography and echocardiography together reveal more high-risk findings than echocardiography alone in the diagnostics of stroke etiology. Cerebrovasc Dis 2013; 35:521-530. 9. Shafqat S, Kelly PJ, Furie KL. Holter monitoring in the diagnosis of stroke mechanism. Intern Med J 2004; 34:305-309. 10. Staszewski J, Brodacki B, Tomczykiewicz K, et al. Strokes in paroxysmal atrial fibrillation have more favorable outcome than in permanent atrial fibrillation. Acta Neurol Scand 2009;119:325-331. 11. Rehman SU, Januzzi JL Jr. Natriuretic peptide testing in clinical medicine. Cardiol Rev 2008;16:240-249. 12. Biteker M, Ozden T, Dayan A, et al. Aortic stiffness and plasma brain natriuretic peptide predicts mortality in acute ischemic stroke. Int J Stroke 2013. In press. 13. Fonseca AC, Matias JS, Pinho e Melo T, et al. N-terminal probrain natriuretic peptide as a biomarker of cardioembolic stroke. Int J Stroke 2011;6:398-403. 14. Fonseca AC, Matias JS, e Melo TP, et al. Time course of NT-proBNP levels after acute ischemic stroke. Acta Neurol Scand 2013;128(4):235-240. 15. Hajsadeghi S, Kashani Amin L, Bakhshandeh H, et al. The diagnostic value of N-terminal pro-brain natriuretic peptide in differentiating cardioembolic ischemic stroke. J Stroke Cerebrovasc Dis 2013;22:554-560.

7 16. Naya T, Yukiiri K, Hosomi N, et al. Brain natriuretic peptide as a surrogate marker for cardioembolic stroke with paroxysmal atrial fibrillation. Cerebrovasc Dis 2008;26:434-440. 17. Okada YMT, Nakamura K, Murata H, et al. The significance of serum levels of pro-brain natriuretic peptide for primary diagnosis in cardioembolic stroke. Int J Stroke 2010;5:182. 18. Rodriguez-Yanez M, Sobrino T, Blanco M, et al. High serum levels of pro-brain natriuretic peptide (pro BNP) identify cardioembolic origin in undetermined stroke. Dis Markers 2009;26:189-195. 19. Sakai K, Shibazaki K, Kimura K, et al. Brain natriuretic peptide as a predictor of cardioembolism in acute ischemic stroke patients: brain natriuretic peptide stroke prospective study. Eur Neurol 2013;69:246-251. 20. Santamarina E, Penalba A, Garcia-Berrocoso T, et al. Biomarker level improves the diagnosis of embolic source in ischemic stroke of unknown origin. J Neurol 2012;259:2538-2545. 21. Shibazaki K, Kimura K, Iguchi Y, et al. Plasma brain natriuretic peptide can be a biological marker to distinguish cardioembolic stroke from other stroke types in acute ischemic stroke. Intern Med 2009;48:259-264. 22. Tamura H, Watanabe T, Nishiyama S, et al. Elevated plasma brain natriuretic peptide levels predict left atrial appendage dysfunction in patients with acute ischemic stroke. J Cardiol 2012;60:126-132. 23. Yukiiri K, Hosomi N, Naya T, et al. Plasma brain natriuretic peptide as a surrogate marker for cardioembolic stroke. BMC Neurol 2008;8:45. 24. Zhixin W, Lianhong Y, Wei H, et al. The value of the use of plasma B-type natriuretic peptide among acute ischemic stroke patients in a Chinese emergency department. Clin Neurol Neurosurg 2013;115(9):1671-1676. 25. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA Statement. Open Med 2009;3:e123-e130. 26. Leeflang MM, Deeks JJ, Gatsonis C, et al. Systematic reviews of diagnostic test accuracy. Ann Intern Med 2008; 149:889-897. 27. Whiting P, Rutjes AW, Reitsma JB, et al. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003;3:25. 28. Reitsma JB, Glas AS, Rutjes AW, et al. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005;58:982-990. 29. Chappell FM, Raab GM, Wardlaw JM. When are summary ROC curves appropriate for diagnostic meta-analyses? Stat Med 2009;28:2653-2668. 30. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ 2003;327:557-560. 31. Egger M, Davey Smith G, Schneider M, et al. Bias in metaanalysis detected by a simple, graphical test. BMJ 1997; 315:629-634. 32. Lovett JK, Coull AJ, Rothwell PM. Early risk of recurrence by subtype of ischemic stroke in population-based incidence studies. Neurology 2004;62:569-573. 33. Donnan GA. The burden of stroke and priorities for stroke research. Int J Stroke 2013;8:397. 34. Marsiliani D, Buccelletti F, Carroccia A, et al. Natriuretic peptides and atrial fibrillation. Eur Rev Med Pharmacol Sci 2010;14:855-860. 35. Hijazi Z, Oldgren J, Andersson U, et al. Cardiac biomarkers are associated with an increased risk of stroke and death in patients with atrial fibrillation: a Randomized Evaluation of

8 Long-term Anticoagulation Therapy (RE-LY) substudy. Circulation 2012;125:1605-1616. 36. Montaner J, Garcia-Berrocoso T, Mendioroz M, et al. Brain natriuretic peptide is associated with worsening and mortality in acute stroke patients but adds no prognostic value to clinical predictors of outcome. Cerebrovasc Dis 2012;34:240-245. 37. Molina CA, Montaner J, Abilleira S, et al. Time course of tissue plasminogen activator-induced recanalization in

H.-L. YANG ET AL. acute cardioembolic stroke: a case-control study. Stroke 2001;32:2821-2827. 38. Montaner J, Perea-Gainza M, Delgado P, et al. Etiologic diagnosis of ischemic stroke subtypes with plasma biomarkers. Stroke 2008;39:2280-2287. 39. Sa T, Sargento-Freitas J, Pinheiro V, et al. CHADS and CHADSVASc scores as predictors of cardioembolic sources in secondary stroke prevention. Rev Port Cardiol 2013;32:373-378.

Predicting cardioembolic stroke with the B-type natriuretic peptide test: a systematic review and meta-analysis.

We performed a systematic review and meta-analysis to evaluate the value of B-type natriuretic peptide (BNP) in differentiating cardioembolic (CE) str...
936KB Sizes 0 Downloads 3 Views