Thrombosis Research 135 (2015) 289–291
Contents lists available at ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Regular Article
NT-proBNP predicts maintenance of sinus rhythm after electrical cardioversion Jonas Andersson a, Mårten Rosenqvist b, Per Tornvall c, Kurt Boman a a b c
Department of Medicine and Geriatrics, Skellefteå County Hospital, S-931 86 Skellefteå, Sweden Karolinska Institutet, Department of Clinical Scienecs, Cardiology Unit, Danderyd University Hospital, Stockholm, Sweden Department of Clinical Science and Education Södersjukhuset, Karolinska Institutet, Stockholm, Sweden S-118 83 Stockholm, Sweden
a r t i c l e
i n f o
Article history: Received 18 August 2014 Received in revised form 16 October 2014 Accepted 14 November 2014 Available online 20 November 2014
a b s t r a c t Background: Atrial fibrillation (AF) is the most common arrhythmia. NT-proBNP is a fragment of the prohormone brain natriuretic peptide. Previous studies indicate that increased levels of NT-proBNP are associated with higher recurrence rates of AF after electrical cardioversion. Our null hypothesis was that NT-proBNP does not predict recurrence of AF after restoration of sinus rhythm. Methods: We performed a hypothesis generating study within a double-blinded, placebo-controlled, randomized, prospective multicentre study of the effects of atorvastatin on recurrence of AF after electrical cardioversion. 199 patients with persistent AF and an indication for cardioversion were included in the present substudy. NT-proBNP was assessed prior to cardioversion. Cardioversion was performed according to local standard clinical practice on an elective outpatient basis. Patients were followed-up one month after cardioversion. Results: 181 patients had a successful cardioversion and 91 of the study group remained in sinus rhythm at day 30. Recurrence of AF was observed in 108 patients at day 30. An optimal cutpoint for NT-proBNP at 500 ng/L predicted recurrence of AF after cardioversion (OR 2.94; 95% CI 1.30-6.63). In multivariate analysis adjusting for age, sex, hypertension, and treatment group strengthened the results (OR 3,56; 95% CI 1,44-8,81). When analysing the ROC curve of NT-proBNP in baseline and atrial fibrillation at day 30 the result was 0.57. Conclusion: NT-proBNP levels are a predictor of recurrence of AF 30 days after cardioversion. ROC curves indicates that the practical value of NT-proBNP for the individual patient is limited. © 2014 Elsevier Ltd. All rights reserved.
Introduction Atrial fibrillation (AF) is the most common arrhythmia with a prevalence of 3% in Sweden and is more common among older adults [1,2]. An initial treatment strategy is to choose between rate or rhythm control, two strategies with similar clinical outcome [3]. This can be done through clinical evaluation including the European Heart Rhythm Association (EHRA) score to evaluate AF related symptoms [4]. The recurrence rate of atrial fibrillation at one year after cardioversion is however as high as 71% to 84% [5]. Improved scoring systems could help us deciding for or against, or plan early follow-up during, rhythm control management. NT-proBNP is an inactive fragment of the prohormone brain natriuretic peptide (BNP) which is primarily produced by the ventricles of the heart [6]. The current main clinical utility of NT-pro BNP is to improve the diagnostic accuracy of heart failure [7]. NT-proBNP levels are also increased during AF possibly through production in the atria [8–11]. Thus NT-proBNP could be a possible biomarker for recurrence of AF after restoration of sinus rhythm. Results from previous studies are conflicting but based upon small studies [12–14] or from patients with lone fibrillation [15]. We present prospective data from, to the best of our knowledge, the largest study with the primary aim to evaluate NT-proBNP as a predictor of recurrence of AF after cardioversion.
http://dx.doi.org/10.1016/j.thromres.2014.11.014 0049-3848/© 2014 Elsevier Ltd. All rights reserved.
The null hypothesis was that NT-proBNP does not predict recurrence of AF after restoration of sinus rhythm. Method We conducted a hypothesis generating study within an investigatorinitiated study where atorvastatin was not statistically superior to placebo with regards to maintaining sinus rhythm after electrical cardioversion [16]. This study, a double-blinded, placebo-controlled, randomized, prospective multicentre study was performed at 10 different clinical centres in Sweden and has previously been described in detail [16]. Between August 2004 and January 2007, a total of 234 patients with persistent AF and an indication for cardioversion were included in the study and of which 199 were sampled and included in the present substudy. Persistent AF was defined as AF with duration of more than 7 days. Patients below 18 years and over 80 years of age were excluded. Patients with paroxysmal AF, atrial flutter, contraindications against atorvastatin, ongoing treatment with lipid-lowering drugs, ongoing treatment with class I or class III antiarrhythmic treatment, myopathy, known liver disease and patients with a previous electrical cardioversion the last 12 month were also excluded. Oral amiodarone were not allowed 6 months before inclusion. Treatment with beta blocking agents,
290
J. Andersson et al. / Thrombosis Research 135 (2015) 289–291
calcium antagonists, diuretics, digoxin, ACE-inhibitors, ARBs, and acetylsalicylic acid were used when clinically indicated. Before randomization a 12-lead electrocardiography and detailed medical history was obtained and basic laboratory analyses were performed. NT-proBNP was assessed at baseline (prior to cardioversion). Medical records and a questionnaire were used to identify hypertension, smoking, diabetes mellitus and heart failure. At inclusion, patients were assigned on an individual basis to atorvastatin (40 mg, two tablets once daily) or placebo, initiated at least 14 days before the elective cardioversion. Cardioversion was performed according to local standard clinical practice on an elective outpatient basis. Patients were followed-up at two days and one month after cardioversion. Patients were instructed to contact the study team if they experienced symptoms suggesting AF between scheduled visits. No patient was lost to follow-up. Blood Sampling and Laboratory Measurements Blood samples (plastic EDTA tubes) for analysis of natriuretic peptides were taken in fasting patients who had rested for twenty minutes. After five minutes the samples were centrifuged for 10 minutes at 4 ° C and then stored frozen at – 70 ° C. NT-proBNP was analyzed with Roche Elecsys proBNP immunoassay [17]. Statistical Analysis Values of NT-proBNP are expressed as median and interquartile range. All other variables are given as mean or as percentage. For statistical comparison of NT-proBNP values, the Mann–Whitney test (2 groups) was applied. Correlation was tested with Spearman´s test. Logistic regression was used to calculate the odds ratio and 95% confidence interval for NT-proBNP as a predictor for recurrence of AF. In the first multivariable model the variables with the strongest correlation to NT-proBNP was added (age and sex, p b 0.001) and hypertension as it is important in the pathogenesis of AF. In the second model the treatment group was added (Atorvastatin or placebo). For all analyses p b 0.05 (two-tailed) was considered statistically significant. All analysis was performed using PASW Statistics 18, release 18.0.3. Ethics The Ethics Committee at Karolinska Institute and the Swedish Medical Product Agency approved this study that also complied with the Declaration of Helsinki. Written informed consent was obtained from all patients. This study was investigator-initiated and the manufacturer of atorvastatin, Pfizer, had no other involvement than providing the study medication, and an unrestricted research grant.
Table 1 Baseline characteristics. At day 30 Number (n) Age Sex M/F Smokers Body mass index Hypertension Heart failure Diabetes IHD NT-pro BNP Medication ACE inhibitors Digitoxin Beta-blockers Calcium channel blockers Successfull cardioversion
total n = 199
sinus n = 91
AF n = 108
(years) n (% men) n (%) (kg/m2) n (%) n (%) n (%) n (%) ng/L
65 ± 9.8 150/49 (75)
65 ± 10.0 69/22 (76)
65 ± 9.7 81/27 (75)
20 (10) 28.7 ± 5.5 90 (45) 19 (10) 18 (9) 10 (5) 712/1167/1670
5 (5) 28.5 ± 4.9 48 (53) 12 (13) 10 (11) 6(7) 552/1100/1575
15 (14) 28.9 ± 6.0 42 (39) 7 (6) 8 (7) 4 (4) 789/1233/1706
n (%) n (%) n (%) n (%)
78 (39) 20 (10) 162 (81) 27 (14)
34 (37) 9 (10) 79 (87) 12 (13)
44 (41) 11 (10) 83 (77) 15 (14)
n (%)
181 (91)
the distribution of NT-proBNP concentrations at baseline among those who remained in sinus rhythm and the group with recurrence of atrial fibrillation at day 30. Recurrence of AF was observed in 108 (54%) patients at day 30. Among patients with the highest NT-proBNP levels, defined as more then 1500 ng/L at baseline, 28 of 67 (42%) had sinus rhythm at day 30 compared to 63 of 132 (48%) of those with NTproBNP less than 1500 ng/L. We tested manually for different cut-off points (data not shown) for NT-proBNP and found that 500 ng/L was optimal. Using this cut-off point in univariate analysis NT-proBNP predicted recurrence of AF after cardioversion (OR 2.94; 95% CI 1.30-6.63). In multivariate analysis adding variables correlating to NT-proBNP strengthened the results (OR 3,56; 95% CI 1,44-8,81). Table 2 shows the univariate analysis and multivariate logistic regression models for NT-proBNP as a predictor of sinus rhythm 30 days after cardioversion. We also tested further adjustments for other important background characteristics but they did not change the results other than marginally (data not presented). When analysing NT-proBNP as continuous variable it did not predict recurrence of AF at day 30 after cardioversion. When analysing the ROC curve of NT-proBNP at baseline and AF at day 30 the result was 0.57 with the optimal NTproBNP cut point 500 ng/L and 0.66 using continuous NT-proBNP.
Results In total, 234 patients were randomized in the study. Of these were 14 excluded due to missing NT-proBNP in baseline, 7 because they had spontaneously returned to sinus rhythm prior to cardioversion and 14 due to missing data on rhythm, or other rhythms then sinus rhythm or AF, at day 30. Thus, 199 patients were eligible for the substudy. Of these 150 were male and 49 female, aged 65 years. Baseline characteristics are presented in Table 1. NT-proBNP levels increased significantly with age (r = 0.39, p = 0.000). Levels were significant higher among women (median 1580 ng/L compared to 1025 ng/L among men, p = 0.000). Other significant correlations were found for heart failure (p = 0.023) and BMI (r = −0.15, p = 0.032). NT-proBNP as a Predictor of Recurrent AF after Cardioversion In total, 181 (91%) had a successful cardioversion and 91 (46%) of the study group remained in sinus rhythm at day 30. Fig. 1 illustrats
Fig. 1. Distribution of NT-proBNP concentrations at baseline among those who remained in sinus rhythm and the group with recurrence of atrial fibrillation at day 30.
J. Andersson et al. / Thrombosis Research 135 (2015) 289–291
291
Table 2 Univariate and multivariate logistic regression analysis for predicting sinus rhythm 30 days after cardioversion. Odds ratio per unit with 95% confidence interval are presented.
NT-proBNP Cutpoint 500 ng/L
Univariate
NT-proBNP, age, sex and hypertenstion
BNP, age, sex, hypertenstion and treatment group
2.94⁎ (1.30-6.63)c
3,55⁎ (1,44-8,77)
3,56⁎ (1,44-8,81)
⁎ denotes p b 0.05.
Discussion
Conclusion
The main finding in our study was that NT-proBNP levels with a cutoff level of 500 ng/L at baseline predicted recurrence of AF 30 days after cardioversion. Previous studies testing this hypothesis among different populations and time-ranges have reported conflicting results [12–15]. Many of these studies have been small but the heterogeneity of atrial fibrillation and its pathology may also contribute to diverging results. A meta-analysis from 2011 by Tang et al. found higher BNP levels in patients with AF recurrence [18] among the ten included studies. Their results also suggested statistically heterogeneous data possibly due to differences between study populations. In line with previous studies of AF populations we found significant correlations between NT-proBNP and age and sex [14]. An analysis of ROC curves from our data shows that NT-proBNP as predictor has a rather large overlap between cases with and without relapse of AF diminishing its use in clinical decision making. Möllman have also calculated ROC curves but with better results than ours even though the study population was notably smaller and among subjects with lone fibrillation [15]. There is also a possibility that subgroups may show better results as our population includes subjects with and without heart failure, medication that may affect NT-proBNP levels and all forms of AF. Excluding heart failure patients did not change our results much. Thus the practical value of NT-proBNP to asses the risk of recurrence for the individual patient is limited. How NT-proBNP can predict recurrence of AF is not well understood but it seems likely that NT-proBNP levels in AF derive from the atria where the main endocrine production resides in the absence of heart failure [19]. Myocyte stretch is the main stimulus for proBNP production, but in heart failure neurohormonal activation and hypoxia contributes [20]. As atrial dilation and heart failure increases the risk of AF the hypothesis seems possible [21–23]. This study contributes to the knowledge of NT-proBNP as a predictor for AF and is to our knowledge the largest so far. We have a growing knowledge of not only NT-proBNP but also other factors that increases the risk of recurrence of atrial fibrillation but still lack a simple and clinical relevant scoring system to predict the risk of recurrence of atrial fibrillation. Such scoring system could be a valuable tool when deciding between rhythm control management in clinical practice. Combining other biochemical markers such as troponin or inflammation markers with BNP data might contribute to refine the the sensitivtty and specificity. NT-proBNP may be a part of such a scoring system as it is a wellknown, affordable and available method. Still, further research is needed. The ROC curves are disappointing in our study. The heterogeneity of AF patients may not only explain the mixed results of available NTproBNP studies of AF but also hamper creation of a single scoring system for recurrence of AF.
NT-proBNP levels is a predictor of recurrence of AF 30 days after cardioversion. Subjects with NT-proBNP above 500 ng/L are 3.6 times more likely to relapse than those below 500 ng/L. The specificity and sensitivity is however disappointing in our heterogeneous group which may limit the clinical utility of this finding.
Limitations This is an explorative substudy and the results should be regarded as hypothesis generating. Periods of asymptomatic AF, which is known to be common, was not detected unless they coincided with the followup after cardioversion. The study population includes participants with and without heart failure and medication that may influence NTproBNP levels. This may hamper interpretation of the results but is on the other hand more related to the everyday clinical practice.
Conflict of Interest Statement The authors declare that there are no conflicts of interest. References [1] Norberg J, Bäckström S, Jansson JH, Johansson L. Estimating the prevalence of atrial fibrillation in a general population using validated electronic health data. Clin Epidemiol 2013;5:475–81. [2] Friberg L, Bergfeldt L. Atrial fibrillation prevalence revisited. J Intern Med 2013;274: 461–8. [3] VanGelder IC, Hagens VE, Bosker HA, Kingma JH, Kamp O, Kingma T, et al. A comparison of rate control and rhythm control in patients with recurrent persistent atrial fibrillation. N Engl J Med 2002;23:1834–40. [4] Kirchhof P, Bax J, Blomstrom-Lundquist C, Calkins H, Camm AJ, Cappato R, et al. Early and comprehensive management of atrial fibrillation: Executive summary of the proceedings from the 2nd AFNET-EHRA consensus conference 'research perspectives in AF'. Eur Heart J 2009;30(24):2969–77c. [5] Lafuente-Lafuente C, Mouly S, Longas-Tejero MA, Bergmann JF. Antiarrhythmics for maintaining sinus rhythm after cardioversion of atrial fibrillation. Cochrane Database Syst Rev 2007;17:CD005049. [6] Daniels LB, Maisel AS. Natriuretic peptides. J Am Coll Cardiol 2007;50:2357–68. [7] Tang WH, Francis GS, Morrow DA, Newby LK, Cannon CP, Jesse RL, et al. National academy of clinical biochemistry laboratory medicine practice guidelines: Clinical utilization of cardiac biomarker testing in heart failure. Circulation 2007;116:99–109. [8] Inoue S, Murakami Y, Sano K, Katoh H, Shimada T. Atrium as a source of brain natriuretic polypeptide in patients with atrial fibrillation. J Card Fail 2000;6:92–6. [9] Knudsen CW, Omland T, Clopton P, Westheim A, Wu AH, Duc P, et al. Impact of atrial fibrillation on the diagnostic performance of b-type natriuretic peptide concentration in dyspneic patients: An analysis from the breathing not properly multinational study. J Am Coll Cardiol 2005;46:838–44. [10] Lee SH, Jung JH, Choi SH, Lee N, Park WJ, Oh DJ, et al. Determinants of brain natriuretic peptide levels in patients with lone atrial fibrillation. Circ J 2006;1:100–4. [11] Wozakowska-Kapłon B. Effect of sinus rhythm restoration on plasma brain natriuretic peptide in patients with atrial fibrillation. Am J Cardiol 2004;12:1555–8. [12] Ari H, Binici S, Ari S, Akkaya M, Koca V, Bozat T, et al. The predictive value of plasma brain natriuretic peptide for the recurrence of atrial fibrillation six months after external cardioversion. Turk Kardiyol Dern Ars 2008;36:456–60. [13] Shin DI, Jaekel K, Schley P, Sause A, Müller M, Fueth R, et al. Plasma levels of nt-probnp in patients with atrial fibrillation before and after electrical cardioversion. Z Kardiol 2005;94:795–800. [14] Tveit A, Seljeflot I, Grundvold I, Abdelnoor M, Arnesen H, Smith P. Candesartan, ntprobnp and recurrence of atrial fibrillation after electrical cardioversion. Int J Cardiol 2009;131:234–9. [15] Möllmann H, Weber M, Elsässer A, Nef H, Dill T, Rixe J, et al. Nt-probnp predicts rhythm stability after cardioversion of lone atrial fibrillation. Circ J 2008;72:921–5. [16] Almroth H, Höglund N, Boman K, Englund A, Jensen S, Kjellman B, et al. Atorvastatin and persistent atrial fibrillation following cardioversion: A randomized placebocontrolled multicentre study. Eur Heart J 2009;30:827–33. http://dx.doi.org/10. 1093/eurheartj/ehp006. [17] Roche. Roche diagnostic corporations 11/12/02. 51 o(k) summary. K022516-elecsys probnp; 2002 [Nov 1 9 2002. Introduction]. [18] Tang Y, Yang H, Qiu J. Relationship between brain natriuretic peptide and recurrence of atrial fibrillation after successful electrical cardioversion: A meta-analysis. J Int Med Res 2011;39:1618–24. [19] Goetze JP, Friis-Hansen L, Rehfeld JF, Nilsson B, Svendsen JH. Atrial secretion of btype natriuretic peptide. Eur Heart J 2006;14:1648–50. [20] Hall C. Nt-probnp: The mechanism behind the marker. J Card Fail 2005(5 Suppl.):81–3. [21] Henry WL, Morganroth J, Pearlman AS, Clark CE, Redwood DR, Itscoitz SB, et al. Relation between echocardiographically determined left atrial size and atrial fibrillation. Circulation 1976;2:273–9. [22] Janse MJ. Why does atrial fibrillation occur? Eur Heart J 1997;18:C12–8. [23] Maisel WH, Stevenson LW. Atrial fibrillation in heart failure: Epidemiology, pathophysiology, and rationale for therapy. Am J Cardiol 2003;6A:2D–8D.
Contents lists available at ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Regular Article
NT-proBNP predicts maintenance of sinus rhythm after electrical cardioversion Jonas Andersson a, Mårten Rosenqvist b, Per Tornvall c, Kurt Boman a a b c
Department of Medicine and Geriatrics, Skellefteå County Hospital, S-931 86 Skellefteå, Sweden Karolinska Institutet, Department of Clinical Scienecs, Cardiology Unit, Danderyd University Hospital, Stockholm, Sweden Department of Clinical Science and Education Södersjukhuset, Karolinska Institutet, Stockholm, Sweden S-118 83 Stockholm, Sweden
a r t i c l e
i n f o
Article history: Received 18 August 2014 Received in revised form 16 October 2014 Accepted 14 November 2014 Available online 20 November 2014
a b s t r a c t Background: Atrial fibrillation (AF) is the most common arrhythmia. NT-proBNP is a fragment of the prohormone brain natriuretic peptide. Previous studies indicate that increased levels of NT-proBNP are associated with higher recurrence rates of AF after electrical cardioversion. Our null hypothesis was that NT-proBNP does not predict recurrence of AF after restoration of sinus rhythm. Methods: We performed a hypothesis generating study within a double-blinded, placebo-controlled, randomized, prospective multicentre study of the effects of atorvastatin on recurrence of AF after electrical cardioversion. 199 patients with persistent AF and an indication for cardioversion were included in the present substudy. NT-proBNP was assessed prior to cardioversion. Cardioversion was performed according to local standard clinical practice on an elective outpatient basis. Patients were followed-up one month after cardioversion. Results: 181 patients had a successful cardioversion and 91 of the study group remained in sinus rhythm at day 30. Recurrence of AF was observed in 108 patients at day 30. An optimal cutpoint for NT-proBNP at 500 ng/L predicted recurrence of AF after cardioversion (OR 2.94; 95% CI 1.30-6.63). In multivariate analysis adjusting for age, sex, hypertension, and treatment group strengthened the results (OR 3,56; 95% CI 1,44-8,81). When analysing the ROC curve of NT-proBNP in baseline and atrial fibrillation at day 30 the result was 0.57. Conclusion: NT-proBNP levels are a predictor of recurrence of AF 30 days after cardioversion. ROC curves indicates that the practical value of NT-proBNP for the individual patient is limited. © 2014 Elsevier Ltd. All rights reserved.
Introduction Atrial fibrillation (AF) is the most common arrhythmia with a prevalence of 3% in Sweden and is more common among older adults [1,2]. An initial treatment strategy is to choose between rate or rhythm control, two strategies with similar clinical outcome [3]. This can be done through clinical evaluation including the European Heart Rhythm Association (EHRA) score to evaluate AF related symptoms [4]. The recurrence rate of atrial fibrillation at one year after cardioversion is however as high as 71% to 84% [5]. Improved scoring systems could help us deciding for or against, or plan early follow-up during, rhythm control management. NT-proBNP is an inactive fragment of the prohormone brain natriuretic peptide (BNP) which is primarily produced by the ventricles of the heart [6]. The current main clinical utility of NT-pro BNP is to improve the diagnostic accuracy of heart failure [7]. NT-proBNP levels are also increased during AF possibly through production in the atria [8–11]. Thus NT-proBNP could be a possible biomarker for recurrence of AF after restoration of sinus rhythm. Results from previous studies are conflicting but based upon small studies [12–14] or from patients with lone fibrillation [15]. We present prospective data from, to the best of our knowledge, the largest study with the primary aim to evaluate NT-proBNP as a predictor of recurrence of AF after cardioversion.
http://dx.doi.org/10.1016/j.thromres.2014.11.014 0049-3848/© 2014 Elsevier Ltd. All rights reserved.
The null hypothesis was that NT-proBNP does not predict recurrence of AF after restoration of sinus rhythm. Method We conducted a hypothesis generating study within an investigatorinitiated study where atorvastatin was not statistically superior to placebo with regards to maintaining sinus rhythm after electrical cardioversion [16]. This study, a double-blinded, placebo-controlled, randomized, prospective multicentre study was performed at 10 different clinical centres in Sweden and has previously been described in detail [16]. Between August 2004 and January 2007, a total of 234 patients with persistent AF and an indication for cardioversion were included in the study and of which 199 were sampled and included in the present substudy. Persistent AF was defined as AF with duration of more than 7 days. Patients below 18 years and over 80 years of age were excluded. Patients with paroxysmal AF, atrial flutter, contraindications against atorvastatin, ongoing treatment with lipid-lowering drugs, ongoing treatment with class I or class III antiarrhythmic treatment, myopathy, known liver disease and patients with a previous electrical cardioversion the last 12 month were also excluded. Oral amiodarone were not allowed 6 months before inclusion. Treatment with beta blocking agents,
290
J. Andersson et al. / Thrombosis Research 135 (2015) 289–291
calcium antagonists, diuretics, digoxin, ACE-inhibitors, ARBs, and acetylsalicylic acid were used when clinically indicated. Before randomization a 12-lead electrocardiography and detailed medical history was obtained and basic laboratory analyses were performed. NT-proBNP was assessed at baseline (prior to cardioversion). Medical records and a questionnaire were used to identify hypertension, smoking, diabetes mellitus and heart failure. At inclusion, patients were assigned on an individual basis to atorvastatin (40 mg, two tablets once daily) or placebo, initiated at least 14 days before the elective cardioversion. Cardioversion was performed according to local standard clinical practice on an elective outpatient basis. Patients were followed-up at two days and one month after cardioversion. Patients were instructed to contact the study team if they experienced symptoms suggesting AF between scheduled visits. No patient was lost to follow-up. Blood Sampling and Laboratory Measurements Blood samples (plastic EDTA tubes) for analysis of natriuretic peptides were taken in fasting patients who had rested for twenty minutes. After five minutes the samples were centrifuged for 10 minutes at 4 ° C and then stored frozen at – 70 ° C. NT-proBNP was analyzed with Roche Elecsys proBNP immunoassay [17]. Statistical Analysis Values of NT-proBNP are expressed as median and interquartile range. All other variables are given as mean or as percentage. For statistical comparison of NT-proBNP values, the Mann–Whitney test (2 groups) was applied. Correlation was tested with Spearman´s test. Logistic regression was used to calculate the odds ratio and 95% confidence interval for NT-proBNP as a predictor for recurrence of AF. In the first multivariable model the variables with the strongest correlation to NT-proBNP was added (age and sex, p b 0.001) and hypertension as it is important in the pathogenesis of AF. In the second model the treatment group was added (Atorvastatin or placebo). For all analyses p b 0.05 (two-tailed) was considered statistically significant. All analysis was performed using PASW Statistics 18, release 18.0.3. Ethics The Ethics Committee at Karolinska Institute and the Swedish Medical Product Agency approved this study that also complied with the Declaration of Helsinki. Written informed consent was obtained from all patients. This study was investigator-initiated and the manufacturer of atorvastatin, Pfizer, had no other involvement than providing the study medication, and an unrestricted research grant.
Table 1 Baseline characteristics. At day 30 Number (n) Age Sex M/F Smokers Body mass index Hypertension Heart failure Diabetes IHD NT-pro BNP Medication ACE inhibitors Digitoxin Beta-blockers Calcium channel blockers Successfull cardioversion
total n = 199
sinus n = 91
AF n = 108
(years) n (% men) n (%) (kg/m2) n (%) n (%) n (%) n (%) ng/L
65 ± 9.8 150/49 (75)
65 ± 10.0 69/22 (76)
65 ± 9.7 81/27 (75)
20 (10) 28.7 ± 5.5 90 (45) 19 (10) 18 (9) 10 (5) 712/1167/1670
5 (5) 28.5 ± 4.9 48 (53) 12 (13) 10 (11) 6(7) 552/1100/1575
15 (14) 28.9 ± 6.0 42 (39) 7 (6) 8 (7) 4 (4) 789/1233/1706
n (%) n (%) n (%) n (%)
78 (39) 20 (10) 162 (81) 27 (14)
34 (37) 9 (10) 79 (87) 12 (13)
44 (41) 11 (10) 83 (77) 15 (14)
n (%)
181 (91)
the distribution of NT-proBNP concentrations at baseline among those who remained in sinus rhythm and the group with recurrence of atrial fibrillation at day 30. Recurrence of AF was observed in 108 (54%) patients at day 30. Among patients with the highest NT-proBNP levels, defined as more then 1500 ng/L at baseline, 28 of 67 (42%) had sinus rhythm at day 30 compared to 63 of 132 (48%) of those with NTproBNP less than 1500 ng/L. We tested manually for different cut-off points (data not shown) for NT-proBNP and found that 500 ng/L was optimal. Using this cut-off point in univariate analysis NT-proBNP predicted recurrence of AF after cardioversion (OR 2.94; 95% CI 1.30-6.63). In multivariate analysis adding variables correlating to NT-proBNP strengthened the results (OR 3,56; 95% CI 1,44-8,81). Table 2 shows the univariate analysis and multivariate logistic regression models for NT-proBNP as a predictor of sinus rhythm 30 days after cardioversion. We also tested further adjustments for other important background characteristics but they did not change the results other than marginally (data not presented). When analysing NT-proBNP as continuous variable it did not predict recurrence of AF at day 30 after cardioversion. When analysing the ROC curve of NT-proBNP at baseline and AF at day 30 the result was 0.57 with the optimal NTproBNP cut point 500 ng/L and 0.66 using continuous NT-proBNP.
Results In total, 234 patients were randomized in the study. Of these were 14 excluded due to missing NT-proBNP in baseline, 7 because they had spontaneously returned to sinus rhythm prior to cardioversion and 14 due to missing data on rhythm, or other rhythms then sinus rhythm or AF, at day 30. Thus, 199 patients were eligible for the substudy. Of these 150 were male and 49 female, aged 65 years. Baseline characteristics are presented in Table 1. NT-proBNP levels increased significantly with age (r = 0.39, p = 0.000). Levels were significant higher among women (median 1580 ng/L compared to 1025 ng/L among men, p = 0.000). Other significant correlations were found for heart failure (p = 0.023) and BMI (r = −0.15, p = 0.032). NT-proBNP as a Predictor of Recurrent AF after Cardioversion In total, 181 (91%) had a successful cardioversion and 91 (46%) of the study group remained in sinus rhythm at day 30. Fig. 1 illustrats
Fig. 1. Distribution of NT-proBNP concentrations at baseline among those who remained in sinus rhythm and the group with recurrence of atrial fibrillation at day 30.
J. Andersson et al. / Thrombosis Research 135 (2015) 289–291
291
Table 2 Univariate and multivariate logistic regression analysis for predicting sinus rhythm 30 days after cardioversion. Odds ratio per unit with 95% confidence interval are presented.
NT-proBNP Cutpoint 500 ng/L
Univariate
NT-proBNP, age, sex and hypertenstion
BNP, age, sex, hypertenstion and treatment group
2.94⁎ (1.30-6.63)c
3,55⁎ (1,44-8,77)
3,56⁎ (1,44-8,81)
⁎ denotes p b 0.05.
Discussion
Conclusion
The main finding in our study was that NT-proBNP levels with a cutoff level of 500 ng/L at baseline predicted recurrence of AF 30 days after cardioversion. Previous studies testing this hypothesis among different populations and time-ranges have reported conflicting results [12–15]. Many of these studies have been small but the heterogeneity of atrial fibrillation and its pathology may also contribute to diverging results. A meta-analysis from 2011 by Tang et al. found higher BNP levels in patients with AF recurrence [18] among the ten included studies. Their results also suggested statistically heterogeneous data possibly due to differences between study populations. In line with previous studies of AF populations we found significant correlations between NT-proBNP and age and sex [14]. An analysis of ROC curves from our data shows that NT-proBNP as predictor has a rather large overlap between cases with and without relapse of AF diminishing its use in clinical decision making. Möllman have also calculated ROC curves but with better results than ours even though the study population was notably smaller and among subjects with lone fibrillation [15]. There is also a possibility that subgroups may show better results as our population includes subjects with and without heart failure, medication that may affect NT-proBNP levels and all forms of AF. Excluding heart failure patients did not change our results much. Thus the practical value of NT-proBNP to asses the risk of recurrence for the individual patient is limited. How NT-proBNP can predict recurrence of AF is not well understood but it seems likely that NT-proBNP levels in AF derive from the atria where the main endocrine production resides in the absence of heart failure [19]. Myocyte stretch is the main stimulus for proBNP production, but in heart failure neurohormonal activation and hypoxia contributes [20]. As atrial dilation and heart failure increases the risk of AF the hypothesis seems possible [21–23]. This study contributes to the knowledge of NT-proBNP as a predictor for AF and is to our knowledge the largest so far. We have a growing knowledge of not only NT-proBNP but also other factors that increases the risk of recurrence of atrial fibrillation but still lack a simple and clinical relevant scoring system to predict the risk of recurrence of atrial fibrillation. Such scoring system could be a valuable tool when deciding between rhythm control management in clinical practice. Combining other biochemical markers such as troponin or inflammation markers with BNP data might contribute to refine the the sensitivtty and specificity. NT-proBNP may be a part of such a scoring system as it is a wellknown, affordable and available method. Still, further research is needed. The ROC curves are disappointing in our study. The heterogeneity of AF patients may not only explain the mixed results of available NTproBNP studies of AF but also hamper creation of a single scoring system for recurrence of AF.
NT-proBNP levels is a predictor of recurrence of AF 30 days after cardioversion. Subjects with NT-proBNP above 500 ng/L are 3.6 times more likely to relapse than those below 500 ng/L. The specificity and sensitivity is however disappointing in our heterogeneous group which may limit the clinical utility of this finding.
Limitations This is an explorative substudy and the results should be regarded as hypothesis generating. Periods of asymptomatic AF, which is known to be common, was not detected unless they coincided with the followup after cardioversion. The study population includes participants with and without heart failure and medication that may influence NTproBNP levels. This may hamper interpretation of the results but is on the other hand more related to the everyday clinical practice.
Conflict of Interest Statement The authors declare that there are no conflicts of interest. References [1] Norberg J, Bäckström S, Jansson JH, Johansson L. Estimating the prevalence of atrial fibrillation in a general population using validated electronic health data. Clin Epidemiol 2013;5:475–81. [2] Friberg L, Bergfeldt L. Atrial fibrillation prevalence revisited. J Intern Med 2013;274: 461–8. [3] VanGelder IC, Hagens VE, Bosker HA, Kingma JH, Kamp O, Kingma T, et al. A comparison of rate control and rhythm control in patients with recurrent persistent atrial fibrillation. N Engl J Med 2002;23:1834–40. [4] Kirchhof P, Bax J, Blomstrom-Lundquist C, Calkins H, Camm AJ, Cappato R, et al. Early and comprehensive management of atrial fibrillation: Executive summary of the proceedings from the 2nd AFNET-EHRA consensus conference 'research perspectives in AF'. Eur Heart J 2009;30(24):2969–77c. [5] Lafuente-Lafuente C, Mouly S, Longas-Tejero MA, Bergmann JF. Antiarrhythmics for maintaining sinus rhythm after cardioversion of atrial fibrillation. Cochrane Database Syst Rev 2007;17:CD005049. [6] Daniels LB, Maisel AS. Natriuretic peptides. J Am Coll Cardiol 2007;50:2357–68. [7] Tang WH, Francis GS, Morrow DA, Newby LK, Cannon CP, Jesse RL, et al. National academy of clinical biochemistry laboratory medicine practice guidelines: Clinical utilization of cardiac biomarker testing in heart failure. Circulation 2007;116:99–109. [8] Inoue S, Murakami Y, Sano K, Katoh H, Shimada T. Atrium as a source of brain natriuretic polypeptide in patients with atrial fibrillation. J Card Fail 2000;6:92–6. [9] Knudsen CW, Omland T, Clopton P, Westheim A, Wu AH, Duc P, et al. Impact of atrial fibrillation on the diagnostic performance of b-type natriuretic peptide concentration in dyspneic patients: An analysis from the breathing not properly multinational study. J Am Coll Cardiol 2005;46:838–44. [10] Lee SH, Jung JH, Choi SH, Lee N, Park WJ, Oh DJ, et al. Determinants of brain natriuretic peptide levels in patients with lone atrial fibrillation. Circ J 2006;1:100–4. [11] Wozakowska-Kapłon B. Effect of sinus rhythm restoration on plasma brain natriuretic peptide in patients with atrial fibrillation. Am J Cardiol 2004;12:1555–8. [12] Ari H, Binici S, Ari S, Akkaya M, Koca V, Bozat T, et al. The predictive value of plasma brain natriuretic peptide for the recurrence of atrial fibrillation six months after external cardioversion. Turk Kardiyol Dern Ars 2008;36:456–60. [13] Shin DI, Jaekel K, Schley P, Sause A, Müller M, Fueth R, et al. Plasma levels of nt-probnp in patients with atrial fibrillation before and after electrical cardioversion. Z Kardiol 2005;94:795–800. [14] Tveit A, Seljeflot I, Grundvold I, Abdelnoor M, Arnesen H, Smith P. Candesartan, ntprobnp and recurrence of atrial fibrillation after electrical cardioversion. Int J Cardiol 2009;131:234–9. [15] Möllmann H, Weber M, Elsässer A, Nef H, Dill T, Rixe J, et al. Nt-probnp predicts rhythm stability after cardioversion of lone atrial fibrillation. Circ J 2008;72:921–5. [16] Almroth H, Höglund N, Boman K, Englund A, Jensen S, Kjellman B, et al. Atorvastatin and persistent atrial fibrillation following cardioversion: A randomized placebocontrolled multicentre study. Eur Heart J 2009;30:827–33. http://dx.doi.org/10. 1093/eurheartj/ehp006. [17] Roche. Roche diagnostic corporations 11/12/02. 51 o(k) summary. K022516-elecsys probnp; 2002 [Nov 1 9 2002. Introduction]. [18] Tang Y, Yang H, Qiu J. Relationship between brain natriuretic peptide and recurrence of atrial fibrillation after successful electrical cardioversion: A meta-analysis. J Int Med Res 2011;39:1618–24. [19] Goetze JP, Friis-Hansen L, Rehfeld JF, Nilsson B, Svendsen JH. Atrial secretion of btype natriuretic peptide. Eur Heart J 2006;14:1648–50. [20] Hall C. Nt-probnp: The mechanism behind the marker. J Card Fail 2005(5 Suppl.):81–3. [21] Henry WL, Morganroth J, Pearlman AS, Clark CE, Redwood DR, Itscoitz SB, et al. Relation between echocardiographically determined left atrial size and atrial fibrillation. Circulation 1976;2:273–9. [22] Janse MJ. Why does atrial fibrillation occur? Eur Heart J 1997;18:C12–8. [23] Maisel WH, Stevenson LW. Atrial fibrillation in heart failure: Epidemiology, pathophysiology, and rationale for therapy. Am J Cardiol 2003;6A:2D–8D.
