Objectives: Although recent studies suggest that inflammation is involved in the pathogenesis of atrial fibrillation (AF), it remains controversial whether it is a consequence or a cause of the arrhythmia.
Design: Prospective study.
Setting: Tertiary referral centre.
Patients and Interventions: In 52 patients with persistent AF lasting >3 months, high-sensitivity C-reactive protein (hs-CRP) was measured before and after electrical cardioversion.
Measurements and Results: All patients were successfully cardioverted to sinus rhythm (SR), but the recurrence rate was 23% at 1 month. Baseline hs-CRP was higher in patients with AF recurrence than in those who remained in SR (0.5 (SD 0.18) mg/dl vs 0.29 (SD 0.13) mg/dl, respectively, p<0.001). Similarly, arrhythmia recurrence was associated with greater left atrial diameters (45.4 (SD 3.3) mm vs 40.7 (SD 3.1) mm, respectively, p<0.001). However, logistic regression analysis showed that hs-CRP was the only independent predictor for AF recurrence (p<0.001). Additionally, patients who were in SR on final evaluation had significantly lower hs-CRP levels than at baseline (0.10 (SD 0.06) mg/dl vs 0.29 (SD 0.13) mg/dl, respectively, p<0.001), while those who experienced AF recurrence had similar values on final and on initial evaluation (0.56 (SD 0.24) mg/dl vs 0.50 (SD 0.18) mg/dl, respectively, p = 0.42).
Conclusion: High levels of hs-CRP are associated with an increased risk of AF recurrence after cardioversion. The restoration and maintenance of SR result in a gradual decrease of hs-CRP while AF recurrence has a different effect, suggesting that inflammation is a consequence, rather than a cause, of AF.
- C-reactive protein
- atrial fibrillation
Statistics from Altmetric.com
Recent studies suggest that inflammation may play a role in AF, since CRP, a sensitive inflammatory index, is associated with the occurrence and the persistence of AF.1–6 If inflammatory processes are involved in the mechanisms underlying AF, the modification of the inflammatory substrate could be the target for several interventions in order to terminate or prevent AF.
However, it remains controversial whether inflammation is a consequence or a cause of AF.6 Since a relationship between inflammation and AF has been suggested, we planned to evaluate the importance of baseline hs-CRP in predicting both successful termination and recurrence of AF after cardioversion in patients with persistent AF. Moreover, to clarify the relationship between AF and inflammation we evaluated the time course of hs-CRP after cardioversion.
The Ethics Committee of our Institution approved the study. This investigation conforms to the principles outlined in the Declaration of Helsinki. Signed informed written consent was obtained from all subjects before their participation in the study.
The study population consisted of 53 patients who suffered from persistent AF, lasting more than 3 months, and had been referred to our department for cardioversion. The arrhythmia was considered persistent when it was documented on sequential 12-lead electrocardiograms, without any intervening periods of SR. The patients were clinically stable during the 4 weeks preceding the beginning of our study and had normal renal function with plasma creatinine values less than 1.0 mg/dl. Exclusion criteria were malignancy, chronic inflammatory disease, chronic obstructive pulmonary disease, connective tissue disease, valvular heart disease, recent infection and surgery within the previous month. Patients who were over 80 years old, or had had acute coronary syndrome or decompensated heart failure during the previous 6 weeks, and those with a left atrial diameter greater than 51 mm were also excluded.
Diltiazem as well as β-blockers were allowed for ventricular rate control while oral amiodarone was initiated 1 month prior to cardioversion and was continued throughout the study in order to improve the likelihood of a successful cardioversion and to maintain SR thereafter.7
Follow-up and data collection
All subjects underwent a comprehensive clinical evaluation, which included medical history, physical examination, routine laboratory tests, thyroid function tests, and chest X-ray, on their entry into the study. Maximal left atrial diameter and left ventricular ejection fraction (LVEF) were also assessed in all participants by 2-dimensional echocardiography, prior to cardioversion.
Blood samples for hs-CRP measurement were drawn 1 month before and immediately prior to sedation for cardioversion. hs-CRP levels were also determined at 1 day and at 2 weeks after cardioversion and at the end of the 1-month follow-up period, to evaluate their post-cardioversion time course. Patients were prospectively followed for recurrence of AF up to 1 month after cardioversion through telephone interviews and weekly clinic visits. When arrhythmia recurrence was suspected it had to be confirmed by a physician’s examination and electrocardiographic documentation.
Electrical conversion to SR was attempted while patients with AF were properly anticoagulated by oral treatment with acenocoumarol which resulted in an international normalised ratio between 2.0 and 3.0 for at least 4 weeks. Anticoagulation had to be continued for at least 1 month after the procedure, regardless of its outcome. Cardioversion was performed in the electrophysiology laboratory under short-lasting general anaesthesia with intravenous propofol, using a biphasic defibrillator with single synchronised shock energy of 200 J in the anteroposterior paddle position. Cardioversion was considered successful if sinus rhythm was achieved for at least 1 minute after cardioversion. All patients remained in the electrophysiology laboratory and their electrocardiogram was monitored continuously (EP Lab, Quinton, Inc., USA), not only throughout the procedure but for at least 1 h after cardioversion. Thereafter, patients in whom sinus rhythm was restored remained for 24 additional hours in the hospital before their discharge.
Blood samples for hs-CRP measurement were obtained from all patients before and after cardioversion, as mentioned above, after 30 minutes of supine rest. Blood samples were immediately placed on ice and centrifuged within 1 h. Plasma samples were stored at −80.0°C until analysis. hs-CRP was assayed by immunonephelometry using a Dade Behring BN II analyser (Dade Behring Inc., Newark, USA), according to the manufacturer’s protocol, with a typical detection limit of 0.0175 mg/dl and a normal reference range of 0.08 to 0.8 mg/dl. The intra-assay and interassay coefficients of variation were <8% in our laboratory. Measurements were performed by personnel blinded to the clinical trial.
Summary descriptive statistics are presented as mean (standard deviation in brackets). Continuous data were tested for normality by the Kolmogorov–Smirnov test. Baseline continuous variables were compared between the relapsed and non-relapsed groups of patients with a two-sided independent sample t test. Categorical variables were compared using Χ2 test or Fisher’s exact test, as appropriate. Associations between continuous variables of interest were assessed with correlation and linear regression techniques. Repeated measures analysis of variance was used to assess changes in the time course of CRP in relapsed and non-relapsed patients. Ninety-five per cent confidence intervals of mean hs-CRP were also computed in order to gain a clearer idea of the magnitude of differences. Forward stepwise logistic regression analysis was used to determine whether baseline CRP and left atrial diameter could be used to provide independent predictive information for relapse. Only the variables which were significantly correlated with AF recurrence in univariate analysis were entered to identify independent predictors. Additionally, age and sex were not included in the logistic regression analysis because of the small size of the sample. p Values less than 5% were considered significant.
The baseline clinical and demographic features of the patients are shown in table 1. The patient population included 37 men and 15 women, aged between 48 and 74 years (61.6 (SD 7.5)), who suffered from persistent AF. All subjects had normal LVEF except five who had mild left ventricular dysfunction.
Clinical outcome and hs-CRP levels
Of the 52 patients with persistent AF who participated in the study all were successfully cardioverted to SR, but immediate reinitiation of arrhythmia occurred in four of them. None of them converted to sinus rhythm, although at least one new effort of CV was attempted. Furthermore, AF recurred within the study period in eight additional patients (total recurrence rate 23%). All patients tolerated the procedures well without experiencing any complication during follow-up.
The clinical characteristics of patients who experienced AF relapse compared with those who remained in SR are listed in table 1. Age, gender, LVEF, medical history and medications did not differ significantly between these two groups. Baseline hs-CRP levels, determined 1 month prior to cardioversion, were significantly higher in patients with AF on final evaluation than in those who remained in SR after successful cardioversion (0.5 (SD 0.18) mg/dl vs 0.29 (SD 0.13) mg/dl, respectively, p<0.001) (fig 1, table 2). In addition, patients in whom AF recurred during the follow-up period had greater left atrial diameters than those who maintained SR (45.4 (SD 3.3) mm vs 40.7 (SD 3.1) mm, respectively, p<0.001). Logistic regression analysis showed that, of baseline hs-CRP and left atrial diameter, hs-CRP was the only independent predictor for AF recurrence within a month after conversion to SR (β = 7.8, 95% confidence interval 2.8 to 12.8, p<0.001) (fig 1, table 2). Moreover, in all patients taken together, we found that there was a positive correlation between the pre-cardioversion hs-CRP levels and left atrial dimension (r = 0.68, p<0.001) (fig 2), while there was a reverse relationship between hs-CRP and LVEF (r = −0.36, p<0.009).
In addition, the restoration and maintenance of SR resulted in a decrease of hs-CRP levels. Thus, patients who were in SR on final evaluation had significantly lower hs-CRP levels than at baseline (0.10 (SD 0.06) mg/dl vs 0.29 (SD 0.13) mg/dl, respectively, p<0.001) (fig 1, table 2), while those who experienced AF recurrence by the end of the follow-up period had similar values of hs-CRP on final and on initial evaluation (0.56 (SD 0.24) mg/dl vs 0.50 (SD 0.18) mg/dl, respectively, p = 0.42) (fig 1, table 2). In the relapsed group hs-CRP was higher throughout the study period than in the patients who remained in SR (fig 1, table 2). Interestingly, hs-CRP levels determined 1 day after CV increased significantly in both groups of patients compared with their baseline values (fig 1, table 2), possibly due to an inflammatory activation which could be attributed to the electrical shock.
According to the results of recent published studies, there is increasing evidence linking inflammatory processes and AF.1–5 Indeed, the frequent occurrence of AF in clinical situations associated with inflammation supports an inflammatory contribution to at least some types of AF.1–6
In a population with persistent, long-lasting, non-valvular AF we found that baseline hs-CRP levels were a predictor of arrhythmia recurrence after cardioversion. In addition, this study is the first to document that the maintenance of SR resulted in a gradual decrease of hs-CRP levels during the post-cardioversion follow-up period, while AF recurrence had an opposite effect. Moreover, our results indicate the presence of systemic inflammation in patients with AF and suggest that hs-CRP could help to specify a subgroup of persistent AF patients who are most likely to experience arrhythmia recurrence after cardioversion, and thus obtain no advantage from such therapy while still being exposed to potential adverse effects.
Inflammation and AF
Inflammation and fibrosis may contribute to the pathogenesis of AF in humans.6 8–15 Actually, inflammatory processes, including healing and repair, could modify the pattern of myocyte apposition, rearranging atrial myocyte connections and altering cell-to-cell interaction and communication, increasing atrial susceptibility to AF.6 8–15
On the other hand, the fibrillating atria are subjected to a continuous, roughly 10-fold increased stimulation, which results in significant calcium accumulation within the atrial myocytes, threatening cell viability.8 16–18 A programmed cell death pathway is commonly activated, leading to apoptotic loss of atrial myocytes, which are typically replaced by fibrous scar.8 17–19 It is possible that such damage to the atrial myocardium could trigger a low-grade inflammatory response that contributes to the maintenance and perpetuation of AF.8 19
In any case, it is unclear at present whether initiation of AF induces inflammatory responses or whether the presence of a pre-existing systemic inflammatory state promotes the initiation and persistence of AF. In our study, we thought to investigate the time course of hs-CRP levels after cardioversion in patients with persistent AF, in an attempt to clarify the association between inflammation and the arrhythmia. Actually, this study is the first to evaluate the time course of hs-CRP levels after cardioversion in patients with long-lasting AF. According to our findings, the restoration of SR resulted in a gradual decrease of hs-CRP levels, which, at the end of the follow-up period, differed significantly from their pre-cardioversion values, whereas hs-CRP levels were similar to their baseline values in the 12 patients with AF recurrence.
If inflammation contributed to the initiation and precipitation of AF, elevated hs-CRP levels would remain unchanged after the restoration of SR, since our only intervention was the successful CV, which has no known anti-inflammatory properties per se. In other words, since elevated hs-CRP levels decreased after the restoration of SR and the termination of arrhythmia and increased again when AF recurred, we deduce that inflammatory processes are a consequence rather than a cause of AF. Similarly, Acevedo et al demonstrated that patients with paroxysmal AF had only a transient elevation of CRP levels, which returned to normal values after conversion to SR.20
Furthermore, elevated hs-CRP could be the consequence of the atrial remodelling that occurs during persistent AF. We found that increased hs-CRP was positively correlated with left atrial diameter and inversely related to LVEF. We also found that both raised hs-CRP and left atrial enlargement were independent risk factors for AF recurrences. It is thus possible that AF is associated with atrial structural changes that may provoke inflammation or may have an inflammatory basis. Hence, AF may induce or be aggravated by inflammation, which in turn may promote the persistence of AF. However, in accordance with our results, inflammation could not be the initial event but must rather be seen as a consequence of AF, although it exacerbates the arrhythmia’s substrate. It is therefore likely that in AF a vicious circle is set in motion, whereby the arrhythmia causes atrial structural remodelling that triggers a low-grade inflammatory response, which in its turn aggravates the AF, and so on. Evidence of an association between inflammation and atrial structural remodelling is provided by two recent studies relating CRP to left atrial size and pre-cardioversion AF duration.21 22
Comparison with other studies
In general, our findings are in agreement with the results of two recently published studies, which reported that baseline hs-CRP levels could predict the outcome of cardioversion of AF.23 24 There are, however, some significant differences.
Our study population was more homogeneous and the patients did not essentially suffer from structural heart disease, since all but five had normal LVEF. All had AF and none had atrial flutter. In addition, the duration of the arrhythmia was longer compared with the aforementioned studies. Given that our patients had AF for more than 3 months, any potential interaction between inflammatory processes, AF and atrial remodelling could be expected to be more extensive. Moreover, cardioversion was performed using a high initial energy in order to avoid any additional attempt and to eliminate the effect of shocks on hs-CRP levels. In their study, Watanabe et al found that there were no significant differences in the hs-CRP levels at baseline, at 1 month after cardioversion and at the end of follow-up, in patients who maintained SR after cardioversion. In other words, the restoration of SR did not result in a decrease of hs-CRP in patients who remained in SR during follow-up.24 In contrast, in view of our results we suggest that the decrease of hs-CRP should be attributed to the restoration of SR, since it occurred only in non-relapsed patients. Finally, unlike the recent study by Cosgrave et al,25 we found that high levels of hs-CRP are associated with an increased risk of AF recurrence after cardioversion. In addition, patients in whom AF recurred had greater left atrial diameters than those who maintained SR, in contrast to the results of Cosgrave et al. Thus, although we agree with the latter authors that inflammatory mechanisms may not be involved in the pathogenesis of AF, our results could indicate that increased hs-CRP may be the consequence of the atrial remodelling that occurs during persistent AF.
Study limitations and clinical implications
Acute phase proteins, such as CRP, are produced by hepatocytes under the influence of cytokines, mainly interleukin and tumour necrosis factor alpha.26 In the present study we could have measured levels of proinflammatory cytokines to strengthen our results further. In addition, another limitation of our study is the lack of inflammatory indexes in patients before the onset of AF. Also, we did not take the duration of AF into consideration in our analysis because of the lack of reliable data on when the arrhythmia first occurred. Although its duration was more than 3 months in all of our patients, data concerning the AF duration previous to the 3-month point were unavailable in certain of them. Furthermore, we could not rule out the occurrence of some asymptomatic, undocumented episodes of AF in patients who were considered to be in SR during the follow-up period, even with the most careful and close monitoring. Finally, the small sample size and the relatively short follow-up period do not allow major conclusions regarding the relationship of systemic inflammation and AF, although our results are clear and consistent and do not appear to have been influenced by these parameters.
Despite these limitations, our findings add to the growing body of evidence linking inflammatory processes and AF pathophysiology. Although it has not yet become definite whether inflammation is an initiator or a perpetuator of AF, both mechanisms may be present, indicating that inflammation is an integral part of the underlying arrhythmic process. In any case, such an association could be expected to have potential therapeutic implications, since anti-inflammatory agents could exert favourable effects, both on patients with known AF and on those who are at high risk of subsequent development. Recently, Dernellis and Panaretou showed that anti-inflammatory treatment could ameliorate AF, since methylprednisolone reduced the risk of recurrent and permanent AF, lowering hs-CRP levels.27 Interestingly, various therapeutic agents have the potential to modify CRP levels, including statins, angiotensin converting enzyme inhibitors, angiotensin receptor blockers, as well as steroids.6
Finally, since CRP could predict cardioversion outcome, measurement of hs-CRP prior to cardioversion may help to define a subgroup of persistent AF patients who are most likely to relapse, obtaining no advantage from such therapy while still being exposed to potential adverse effects.
In a population with persistent AF we found that baseline hs-CRP is a useful predictor of AF recurrence after cardioversion. Additionally, maintenance of SR results in a gradual decrease of hs-CRP levels while AF recurrence has a different effect, suggesting that inflammation is a consequence, rather than a cause, of AF.
However, further investigation and randomised trials are needed to elucidate the exact role of inflammation in AF, to evaluate the clinical use of hs-CRP as an arrhythmia risk marker in different subgroups of patients and in the general population, and to clarify whether anti-inflammatory interventions are beneficial in the setting of AF.
Competing interests: None.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.