Persistent lone atrial fibrillation: Clinicopathologic study of 19 cases Domenico Corradi, MD,* Sergio Callegari, MD,† Laura Manotti, MD,* David Ferrara, MD,‡ Matteo Goldoni, PhD,§ Rossella Alinovi, BsSc,§ Silvana Pinelli, BsSc,§ Paola Mozzoni, BsSc, PhD,§║ Roberta Andreoli, PhD,§║ Angeliki Asimaki, PhD,¶ Alberto Pozzoli, MD,‡ Gabriella Becchi, PhD,* Antonio Mutti, MD,§ Stefano Benussi, MD, PhD,‡ Jeffrey E. Saffitz, MD, PhD,¶ Ottavio Alfieri, MD‡ From the *Department of Biomedical, Biotechnological, and Translational Sciences (S.Bi.Bi.T.), Unit of Pathology, University of Parma, Parma, Italy, †Division of Cardiology, Vaio Hospital, Fidenza, Italy, ‡ Cardiothoracic Surgery Unit, Department of Cardiology, San Raffaele University Hospital, Milan, Italy, § Department of Clinical and Experimental Medicine, Laboratory of Industrial Toxicology, University of Parma, Parma, Italy, ║Italian Workers’ Compensation Authority (INAIL) Research Center at the University of Parma, Parma, Italy, and ¶Department of Pathology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts. BACKGROUND The extent to which atrial myocardium is remodeled in patients with persistent lone atrial fibrillation (LAF) is largely unknown. OBJECTIVE The purpose of this study was to perform a clinicopathologic investigation in patients with persistent LAF. METHODS We characterized structural and molecular remodeling in atrial biopsies from 19 patients (17 males, mean age 49 years) with persistent (47 days; n ¼ 8) or long-lasting persistent (41 year; n ¼ 11) LAF who underwent surgical ablation. Atrial tissue from 15 autopsy samples without clinicopathologic evidence of heart disease served as controls. RESULTS Morphometric analysis showed cardiomyocyte hypertrophy and greater amounts of myolytic damage and interstitial fibrosis in persistent LAF patients compared to controls (P o.0001). Atrial tissue levels of heme oxygenase-1 and 3nitrotyrosine were increased in persistent LAF patients (P o.001), consistent with oxidative stress. Levels of superoxide dismutase-2, interleukin-8, interleukin-10, tumor necrosis factor-α, and thiobarbituric acid reactive substance were greater in controls than in persistent LAF patients. Immunoreactive signal for connexin43 was reduced more frequently in persistent LAF patients than controls. There was no correlation between features of structural or molecular
Introduction “Lone” atrial fibrillation (LAF), also known as “idiopathic” atrial fibrillation (AF), is an ill-defined clinical entity Drs. Saffitz and Alfieri are co-senior authors. Address reprint requests and correspondence: Dr. Domenico Corradi, Department of Biomedical, Biotechnological, and Translational Sciences (S.Bi.Bi.T.), Unit of Pathology, University of Parma, Via Gramsci 14, 43126 Parma, Italy. E-mail address: [email protected].
1547-5271/$-see front matter B 2014 Heart Rhythm Society. All rights reserved.
remodeling and clinical parameters, including persistent LAF duration. CONCLUSION In persistent LAF patients, the atria are modified by structural remodeling and molecular changes of oxidative stress. Tissue changes in persistent LAF appear to occur early after its onset and are qualitatively no different than those observed in patients with atrial fibrillation related to conventional risk factors. These findings suggest that different types of atrial fibrillation are associated with the same spectrum of tissue lesions. Early intervention to restore sinus rhythm in persistent LAF patients may prevent irreversible tissue change, especially interstitial fibrosis. KEYWORDS Lone atrial fibrillation; Remodeling; Structure; Atrial cardiomyocyte; Pathology ABBREVIATIONS AF ¼ atrial fibrillation; Cx43 ¼ connexin43; HO-1 ¼ heme oxygenase-1; IL ¼ interleukin; LAF ¼ lone atrial fibrillation; NYHA ¼ New York Heart Association; PAS ¼ periodic acid Schiff; SOD-2 ¼ superoxide dismutase-2; TBARS ¼ thiobarbituric acid reactive substance; TNF-α ¼ tumor necrosis factor-α (Heart Rhythm 2014;0:1–9) rights reserved.
I
2014 Heart Rhythm Society. All
characterized mainly by the absence of any apparent explanation or underlying etiology, especially associated cardiopulmonary disorders including hypertension.1 Over the last few years, the number of patients with true LAF has decreased as new epidemiologic associations with AF have emerged (e.g., sleep apnea syndrome, obesity, high coffee or alcohol consumption, endurance sports activities, gene mutations).2 The reported prevalence of LAF has ranged widely between 2% and 30%, presumably reflecting the ambiguous
http://dx.doi.org/10.1016/j.hrthm.2014.02.008
2
Heart Rhythm, Vol 0, No 0, Month 2014
definition and uncertain diagnostic criteria for this supraventricular arrhythmia.3 However, new and more accurate clinical tests, as well as the discovery of new etiologies, have progressively reduced the number of LAF cases with no known cause.4 Patients with LAF do not represent a homogeneous group in terms of cardiovascular and thromboembolic risks. Instead, these patients should be divided into low-risk and high-risk categories, based on the chronicity of their arrhythmia and/or left atrial size, and treated accordingly.5,6 In this regard, an emerging parameter for a more accurate and individualized prognostic definition of LAF involves noninvasive assessment of atrial myocardial structural remodeling.7 Of the various histopathologic features of structural remodeling in AF substrates, myocardial fibrosis seems to be one of the most important, especially in view of its apparent structural stability and negligible regression after restoration of sinus rhythm.8,9 An interesting subset of LAF patients consists of cases that have progressed from paroxysmal to persistent AF in the absence of an underlying associated cardiopulmonary disease.10 To the best of our knowledge, no clinicopathologic studies on such patients have been reported in the English language literature. A thorough understanding of the structural and molecular state characterizing the LAF atrial myocardium certainly could represent the basis for future investigations focusing on structural remodeling prevention in these individuals. Based on recent advances regarding morphologic changes in AF and our own experience in analyzing the histopathology of AF atrial samples, the purpose of this retrospective investigation was to characterize the histopathology of tissue remodeling, quantify expression of molecular markers of oxidative stress and inflammation in left and right atrial endomyocardial biopsies, and correlate these changes with clinical characteristics in patients with persistent LAF undergoing surgical radiofrequency ablation.8,11–14
biopsies were obtained before any new radiofrequency ablation lines were produced through the atrial myocardium.12 Control autopsy tissues were sampled in the same way as those obtained surgically. Each left and right atrial sample was immediately divided into 2 fragments: 1 was fixed in 10% buffered formalin for 24 hours for histopathologic evaluations, and the other was frozen in liquid nitrogen and stored at –801C for subsequent molecular analyses.
Methods
Patient population and controls
Patient population and controls We studied 19 consecutive cases of LAF whose current arrhythmic episode lasted 47 days (“persistent AF” according to the current AF classification). Among these cases, patients whose present AF occurrence lasted Z1 year were further classified as having “long-lasting persistent AF.”1,3 As controls, we used 15 consecutive autopsied hearts from age-matched patients. The study was approved by the San Raffaele Hospital Institutional Review Board. All study patients gave written informed consent.
Atrial myocardial samples In each patient, myocardial samples (about 5 5 2 mm) were excised from the left atrial posterior wall (adjacent to the atriotomy line) and the right atrial free wall (between the 2 caval vein outlets).12 All of the biopsy sites were far from any previous transcatheter ablation lines. Furthermore,
Other methods Methods for histologic, morphometric, and immunofluorescence studies, enzyme-linked immunosorbent assay, and measurements of thiobarbituric acid reactive substance (TBARS), 3-nitrotyrosine levels, and heme oxygenase-1 (HO-1) mRNA are detailed in the Online Supplementary Data.
Statistical analysis Normality of data was assessed by the Shapiro-Wilks test. If there was no evidence against normality, data are expressed as mean ⫾ SD. Comparisons were made using independent (AF patients vs controls) or repeated measures (left atrium vs right atrium comparisons within each single patient or control case) Student t tests. When there were significant deviations from normality, data are expressed as median (with interquartile range), and nonparametric tests were performed (Mann-Whitney U test for independent samples and Wilcoxon test for repeated measures). The Pearson correlation test was applied when no outliers were present. Otherwise, the Spearman correlation test was used. Differences in connexin43 (Cx43) immunoreactive signal distribution were tested using the χ2 test. P o.05 was considered significant. Statistical analysis was performed using SPSS (version 21, IBM, Armonk, NY).
Results Nineteen patients (7.6% of the total AF patients undergoing surgical ablation at San Raffaele Hospital in the same period) fulfilled criteria for the diagnosis of persistent LAF and were enrolled in this study. According to their current AF episode duration, 11 patients (58%) had long-lasting persistent AF.3 General clinical data and preoperative echocardiographic measurements of the LAF patients are listed in Table 1. The average age of the LAF patients was 49 years. All but two were younger than 60 years, consistent with previous studies showing younger age of subjects with LAF than those with AF related to other conditions. There was a striking male predominance (17/19 cases [90%]), as also previously reported for LAF.4 Echocardiographic measures (Table 1) all were within normal limits except for left and right atrial volumes, which were above the reference values.15 Preoperatively, 5 patients had asymptomatic AF (New York Heart Association [NYHA] class I), whereas 14 had AF with palpitations (and on this basis only were considered as being
Corradi et al Table 1
Lone Atrial Fibrillation
3
General patient characteristics and echocardiographic data
No. of cases Age (y) Gender (M/F) Body mass index (kg/m2) BSA (m2) Total heart weight (g) Left atrial volume/BSA (mL/m2)* Right atrial volume/BSA (mL/m2) Left ventricular end-diastolic diameter (mm) left ventricular end-systolic diameter (mm) Left ventricular end-diastolic volume (mL) Left ventricular end-systolic volume (mL) Left ventricular ejection fraction (%) Interventricular septum thickness (mm) Left ventricular posterior wall thickness (mm) Pulmonary artery systolic pressure (mm Hg) Total LAF disease duration (mo) Current LAF episode duration (mo) No. of previous transcatheter ablation procedures (0/1/2/3/4/5) New York Heart Association functional class (I/II/III/IV)
LAF
Controls
P value
19 49 ⫾ 9 (31–66) 17/2 27 ⫾ 3 (18–29) 2.00 ⫾ 0.22 (1.44–2.40)
15 48 ⫾ 9 (33–60) 12/3 26 ⫾ 3 (16–29) 1.95 ⫾ 0.12 (1.38–2.05) 341 ⫾ 40 (270–390)
NS NS NS NS
42 ⫾ 13 (24–77) 34 ⫾ 15 (22–67) 50 ⫾ 4 (39–5) 29 ⫾ 3 (24–35) 95 ⫾ 10 (62–110) 40 ⫾ 6 (25–52) 58 ⫾ 4 (53–68) 10 ⫾ 1 (9–11) 9 ⫾ 1 (8–10) 25 ⫾ 4 (20–30) 91 ⫾ 69 (24–240) 17 ⫾ 19 (2–84) 3/4/4/4/3/1 5/14/0/0
Data are expressed as frequency or mean ⫾ SD (range). BSA ¼ body surface area; LAF ¼ lone atrial fibrillation. * Left atrial volume calculated using biplane Simpson method.
in NYHA class II). All patients underwent the Maze III surgical ablation procedure (13 through minithoracotomy and 6 through sternotomy). Sixteen patients (84%) had previously undergone a variable number (from 1–5, Table 1) of unsuccessful (nonsurgical) catheter ablations. After undergoing surgical AF ablation, all but 1 patient reverted to sinus rhythm. The preoperative pharmacologic treatments are listed in Table 2. All of the patients ate a standard Mediterranean-type diet. In accordance with their NYHA functional classification, all subjects performed mildto-moderate physical activity. General data regarding the control (autopsy) population are also listed in Table 1. Causes of death were traumatic injuries in 12 and nontraumatic cerebral hemorrhage in 3. Prior clinical measurements of atrial volume were not available for the autopsied controls. However, careful evaluation by experienced cardiopathologists indicated that atrial sizes in all control subjects fell within the normal range as previously described.16 Table 2
Preoperative treatment n (%)
Acetylsalicylic acid Allopurinol Amiodarone Angiotensin-converting enzyme inhibitor Angiotensin receptor antagonist Beta-blocker Calcium channel blocker Gastric acid pump inhibitor Oral anticoagulant Statin Data are expressed as no. of patients (percentage).
1 (5) 1 (5) 1 (5) 1 (5) 7 (37) 8 (42) 2 (10) 2 (10) 14 (74) 1 (5)
General histopathology and morphometric evaluation Atrial myocardium from patients with LAF showed significant pathologic changes consisting of hypertrophy, myocytolytic degeneration and glycogen accumulation in cardiomyocytes, remodeling of cardiomyocyte bundle orientation, and accumulation of interstitial and perivascular fibrosis. Representative images are shown in Figure 1, and quantitative morphometric data are given in Table 3. Atrial cardiomyocytes were sometimes arranged in irregular bundles (Figure 1A) compared to controls (Figure 1B). Myocytolytic degeneration was characterized by perinuclear loss of sarcomeres appearing as a clear vacuole extending eccentrically into the adjacent sarcoplasm (Figure 1C). These vacuoles were inconstantly filled with glycogen (seen as reddish material in the periodic acid–Schiff [PAS]-stained histologic sections in Figure 1D). Other cardiomyocytes without apparent loss of sarcomeres also showed moderately increased punctate sarcoplasmic PAS positivity, presumably representing intermyofibrillar glycogen accumulation secondary to cell metabolic changes. Atrial myocardial architecture in AF samples was typically modified by pathologic fibrosis between muscle cells (interstitial fibrosis; Figure 1E) and surrounding the adventitia of small intramural blood vessels (perivascular fibrosis; Figure 1F). In some cases, the subendocardial layer was thickened as a result of mild fibroelastosis. The pattern of fibrosis observed in AF patients was typical of what has been reported in previous studies of tissue remodeling in AF and was distinctly different than the dense local fibrosis resulting from previous catheter ablation procedures. Quantitative analysis of structural remodeling demonstrated significant differences between tissue features from
4
Heart Rhythm, Vol 0, No 0, Month 2014
Figure 1 Representative examples of left atrial structural remodeling in lone atrial fibrillation (LAF). Atria in LAF (A vs controls in B) show changes in cardiomyocyte size and orientation. Some cardiomyocytes show myolysis with characteristic perinuclear loss of sarcomeres (arrowheads in C) and glycogen accumulation (arrowhead in D). Interstitial fibrosis (arrowheads in A and E) and perivascular fibrosis (arrowheads in F) reflect collagen deposition around cardiomyocytes and intramural coronary branches. Bar ¼ 200 μm in A, B, E, and F; bar ¼ 100 μm in C and D.
Table 3
Morphometry LAF
Controls
LA
RA
LA †
RA
2.3 ⫾ 2.2 (0.9–10.1) 0 0 Interstitial fibrosis (%) 4.6 ⫾ 3.9 (1.1–16.9) 2.8 ⫾ 3.3 (0–10.5)† 0.5 ⫾ 0.1 (0–2.2) 0.4 ⫾ 0.2 (0–1.4) Cardiomyocyte myocytolysis (%) 5.1 ⫾ 5.6 (0–20.0)† Cardiomyocyte transverse 16.6 ⫾ 2.7 (12.9–20.6)† 15.5 ⫾ 2.6 (10.9–20.1)† 10.9 ⫾ 0.5 (10.4–13.4) 10.3 ⫾ 0.8 (10.2–13.5) diameter (mm) *†
Data are expressed as average ⫾ SD (range). LA ¼ left atrium; LAF ¼ lone atrial fibrillation; RA ¼ right atrium. * P o.05 (LA vs RA, in patients). † P o.0001 (patients vs controls).
Corradi et al
Lone Atrial Fibrillation
AF patients vs controls for all 3 parameters of tissue remodeling (myocytolysis, hypertrophy and fibrosis; Table 3). Controls showed no fibrosis and virtually no myocytolytic degeneration. PAS-positive material in controls consisted of minute particles distributed homogeneously throughout the sarcoplasm. Changes in cell and tissue structure generally appeared to be more severe in the left vs the right atrium in patients with AF, although this difference was significantly different only for the amount of interstitial fibrosis (Table 3). Cardiomyocyte transverse diameters showed statistically significant correlation between left atrium and right atrium (Pearson test, r ¼ 0.61, P o.05). There were no correlations between any of the histopathologic parameters and the number of previous transcatheter ablations. There was also no correlation between the amount of atrial fibrosis vs patient’s age or AF duration.
Connexin43 pattern Strong immunoreactive signal for Cx43 was seen in 16 of 19 control atrial cardiomyocytes both at intercalated disk regions and along the lateral sarcolemma (Figure 2). In the
5 remaining 3 cases, the amount of signal appeared to be reduced. By contrast, Cx43 signal was diminished in 11 of 19 of the LAF cases (P o.05 compared with controls; Figure 2). Signal was abundant at intercalated disks in samples from 7 AF cases and lateralized in 1 case. Reduced Cx43 signal occurred equally in both atria in 8 of the 11 cases of LAF with reduced signal. Signal was reduced in the left atrium but normal in the right atrium in 1 case and was reduced in the right atrium but normal in the left atrium in 2 cases.
Molecular markers of oxidative stress and inflammation Quantitative measurements of tissue markers of oxidative stress and inflammation are given in Table 4. Expression of the oxidative stress-related protein HO-1 was significantly increased in the left and right atrium (10- and 37-fold, respectively) in LAF patients compared to controls. Increased immunoreactive signal for HO-1 was also readily apparent in atrial cardiomyocytes from LAF patients compared to controls (Figure 3), whereas other markers of oxidative stress or inflammation showed no apparent
Figure 2 Connexin43 (Cx43) distribution in representative cases of lone atrial fibrillation (LAF) and controls. Controls showed strong immunoreactive signal at cell–cell junctions and in some cases at lateral membranes. In both LAF samples, Cx43 signal is significantly depressed. Bar ¼ 40 μm.
6 Table 4
Heart Rhythm, Vol 0, No 0, Month 2014 Proteins, thiobarbituric acid reactive substance, and 3-nytrotyrosine molecular measurements LAF
Controls
LA HO-1 (μg/g proteins) SOD-1 (μg/g proteins) SOD-2 (μg/g proteins) IL-6 (ng/g proteins) IL-8 (ng/g proteins) IL-10 (ng/g proteins) TNF-α (ng/g proteins) TBARS (nmol/g proteins) 3-Nitrotyrosine (mmol/mol tyrosine)
RA *
8.3 (3.0-19.8) 4.6 ⫾ 2.3 0.6 (0.3-1.1)† 26.4 (14.3-41.2) 25.8 ⫾ 16.3† 1.7 (1.1-5.4)† 25.3 ⫾ 16.8 16.5 (6.9-26.1)† 62.9 (46.2-89.5)*
*
14.7 (2.6-24.7) 4.7 ⫾ 2.9 1.2 (0.5-4.6) 41.6 (20.4-75.5) 36.1 ⫾ 27.4 4.3 (2.9-9.2) 37.3 ⫾ 26.3† 19.3 (11.0-79.9) 73.8 (50.4-188.4)*
LA
RA
0.8 (0.6-4.0) 5.2 ⫾ 3.2 1.7 (0.4-3.1) 50.3 (29.0-75.4) 47.7 ⫾ 26.8 5.0 (3.2-8.6) 31.7 ⫾ 17.8 36.1 (23.0-46.3) 28.8 (18.2-39.6) 22.4
0.4 (0.2-1.3) 3.5 ⫾ 1.3 0.7 (0.4-1.1) 40.6 (28.8-66.3) 40.5 ⫾ 27.5 3.6 (2.4-9.5) 19.1 ⫾ 9.6 26.2 (18.4-38.4) (17.0-27.5)
* ¼ p o 0.001, † ¼ p o 0.05. Data are expressed as average ⫾ SD or median value (interquartile range). HO ¼ heme oxygenase-1; IL ¼ interleukin; LA ¼ left atrium; LAF ¼ lone atrial fibrillation; RA ¼ right atrium; SOD ¼ superoxide dismutase; TBARS ¼ thiobarbituric acid reactive substance; TNF-α ¼ tumor necrosis factor-α.
differences in immunostaining compared with controls (data not shown). HO-1 mRNA was detected in both atria of patients with LAF in roughly equal amounts (0.47 ⫾ 0.33 vs 0.48 ⫾ 0.27 arbitrary units for left and right atria, respectively; P ¼ NS). Increased levels of HO-1 protein were associated with significantly greater levels (2- to 3fold) of the oxidative stress marker 3-nitrotyrosine in LAF tissue compared with controls (Table 4). Otherwise, left atrial tissue from LAF patients contained lower levels of superoxide dismutase-2 (SOD-2), interleukin (IL)-8, IL-10, and TBARS than the corresponding left atrial control tissue. Right atrial tissue from LAF patients contained greater amounts of tumor necrosis factor-α (TNF-α) than the corresponding right atrial control tissue (Table 4). The magnitude of these differences was not nearly as great, however, as that seen for HO-1 and 3nitrotyrosine.
Discussion In the present study, we found the same type of structural and molecular remodeling in the atria of patients with LAF as previously described in patients with more conventional disease-associated AF, thus suggesting that the different types of AF may be associated with the same continuum of lesions. The definition of LAF is still based largely on the apparent absence of other disorders that might account for “secondary” AF.2 Therefore, the diagnosis depends on the accuracy of clinical evaluation. Advances in such evaluations and identification of additional risk factors associated with AF have progressively diminished the fraction of cases that initially would have been labeled as “lone,”2 such that some now regard the designation of lone/idiopathic AF as obsolete.17 Nevertheless, the effects of the arrhythmia itself on atrial remodeling may be greater in LAF than in AF
Figure 3 Heme oxygenase-1 (HO-1) immunoreactive signal (green) is located in the cardiomyocyte sarcoplasm, identified by red staining for α-sarcomeric actin (SA). HO-1 signal intensity is much greater in the pathologic atrium than in the control (C). Yellow color indicates green and red signal superimposition. Bar ¼ 40 μm.
Corradi et al
Lone Atrial Fibrillation
associated with conventional risk factors (e.g., mitral valve disease) in which structural and molecular remodeling likely precede onset of AF. In this regard, pacing-induced AF in animals may be a better model of LAF than AF associated with usual risk factors.8 Age 460 years and increased atrial volume have generally been regarded as exclusion criteria for LAF, but recent studies have identified cases of AF without any evident underlying process in patients older than 60 years, suggesting that this age cutoff may be inappropriate.4,18 In the present study, we did not exclude patients with atrial enlargement in the absence of other indications of cardiovascular disease. In such cases, atrial enlargement may be the result rather than the cause of AF,10 although at some point in the natural history atrial enlargement likely is both a cause and a consequence of the arrhythmia. Structurally, the atrial myocardial architecture of our patient population had been altered by changes in cardiomyocytes and the interstitium, including increased cardiomyocyte size, myolysis, and interstitial fibrosis. All of these modifications were significantly greater in the LAF patients than in the control population but were less severe than those seen in our previous studies of AF associated with mitral valve disease. For example, the proportion of left and right atrial cardiomyocytes showing myolytic degeneration was 5.1% and 4.6%, respectively, in LAF compared with 15.1% to 20.8% and 7.16% to 10.4%, respectively, in non-lone forms of AF. Hypertrophy of atrial cardiomyocytes in LAF was only slightly less than that seen in valvular AF (16.6 vs 18.6–19 μm, respectively).11–14 The left atrium showed more interstitial fibrosis than the right atrium in LAF (4.6 ⫾ 3.9 vs 2.3 ⫾ 2.2, respectively, P o.05), perhaps related to the different hemodynamic loads of these 2 atrial chambers in AF.19 On the whole, these data confirm the fact that changes in AF structural remodeling are essentially nonspecific and suggest that, in terms of myocardial architectural distortion, the various forms of AF should be placed in a spectrum of disease where the specific effects of the arrhythmia (including its duration), associated diseases, and aging should be considered in addition to other factors. We found a consistent lack of correlation between AF duration and structural remodeling parameters in our previous studies on valvular AF (even after combining the different cases in a single patient population, unpublished data).11–14 The large variability in structural remodeling as a manifestation of myocardial injury likely is related to many factors, including genetics, therapy, and aging in addition to the usual risk factors implicated in AF initiation and/or perpetuation). It is unlikely that structural remodeling of atrial myocardium develops at the same rate over the natural history of AF. In fact, it has been demonstrated experimentally that histopathologic changes in develop progressively over an interval of 4 months of pacing-induced AF, by which time structural remodeling is fully established.20 Thus, it appears that the atria undergo structural remodeling early after the onset of AF, which might explain the consistent lack
7 of correlation between the severity of structural remodeling and the duration of AF in patients whose arrhythmia has occurred for at least a few months. Lastly, AF duration usually is calculated from the first arrhythmic episode, but the number of episodes over the clinical history of AF may be highly variable and often is unknown. The actual disease interval may vary considerably in patients with the same apparent AF duration. In contrast to the study by Frustaci et al,21 we did not find inflammatory infiltrates in myocardial samples from LAF patients. However, the patient populations were not identical, with paroxysmal AF in the previous investigation and persistent AF in the present study. We cannot definitively rule out the possibility that the fibrosis we observed could have been caused by previous myocarditis, and resulting substrate could have played a role in inducing and/or perpetuating the arrhythmia. Changes in the amount of immunoreactive signal for connexins and redistribution of signal from intercalated disk regions to the lateral margins of atrial cardiomyocytes have been described in AF.22,23 In our study, the amount of Cx43 signal was reduced in significantly more AF samples than controls, providing at least some basis for speculating that reduced coupling may have contributed to AF initiation and/ or maintenance. Because the number of LAF cases was limited, it was not possible to correlate changes in Cx43 signal with AF duration, left atrial dimensions, or the other features of structural and molecular remodeling analyzed in this study. Oxidative stress, in which reactive oxygen species damage both structural and functional proteins in cardiomyocytes, has been implicated in the pathogenesis of AF.8 In this study, we observed significantly increased expression of the protective enzyme HO-1 in LAF subjects compared to controls as well as higher levels of the oxidative stress marker 3-nitrotyrosine, which is formed when reactive nitrogen species (e.g., peroxynitrite, nitryl chloride) interact with the aromatic ring of tyrosine in soluble amino acids and in proteins.24 Previously, we reported increased HO-1 expression (greater in the left atrial posterior wall that in the left atrial appendage) in 2 different patient populations with valvular AF.13,14 HO-1 expression has previously been shown to confer protection in various models of cardiovascular injury or disease, including hypertension, atherosclerosis, myocardial infarction, and allograft rejection.25 Further and more focused investigations are necessary to test whether HO-1 may serve as a predictive marker in AF. In the present study, HO-1 expression was not associated with a parallel increase in SOD-2 or lipid peroxidation (as determined by measurement of TBARS) in the left atrial myocardium. In fact, TBARS levels were significantly decreased compared to controls. This might reflect the activation of antioxidative stress-protective mechanisms, as observed in in vitro models.26 The significantly increased levels of 3-nitrotyrosine also suggest greater atrial nitric oxide levels and ongoing nitric oxide–dependent reactive species-induced oxidative stress.27 Increased myocardial
8 3-nitrotyrosine has previously been reported in AF associated with mitral valve disease.28 Inflammatory markers such as C-reactive protein, TNF-α, IL-2, IL-6, and IL-8 are thought to play a role in initiating and perpetuating AF and AF-related thrombosis through mechanisms involving endothelial damage/dysfunction and platelet activation.29 Whether AF is the cause or a consequence of inflammation is not known, but it is likely that both relationships occur. Inflammation may be a potent trigger of AF, but at the same time AF seems to promote an inflammatory and prothrombotic state.8,29,30 Only limited data are available on the systemic or local nature of inflammation in AF. Liuba et al31 found higher plasma levels of IL-8 in femoral vein, right atrial, and coronary sinus blood, but not in pulmonary vein samples, consistent with a systemic source of inflammation. Marcus et al30 measured higher C-reactive protein levels in the left atrium than in the corresponding coronary sinus and suggested that transcardiac cytokine gradients in AF arise by sequestration of inflammatory cytokines in the heart. The results of the present study do not indicate increased local production/ accumulation of atrial myocardial IL-6, IL-8, IL-10, and TNF-α in LAF patients.
Heart Rhythm, Vol 0, No 0, Month 2014 high stability of collagen fibers, preventing or limiting interstitial fibrosis appears to be an important way to mitigate AF. In this regard, interstitial fibrosis was the only component of structural remodeling that had not regressed 4 months after restoration of sinus rhythm in the goal model of pacinginduced AF.9 Clinical and experimental evidence suggests that interstitial fibrosis and other components of atrial structural remodeling interact with AF in a vicious cycle in which they act as both a cause and a consequence of each other.8 In this view, progression of the abnormal substrate would play a part in the transition from paroxysmal to persistent AF. In view of limitations associated with antiarrhythmic drugs and extensive ablative procedure and with increasing evidence of ongoing oxidative stress, attention has turned to the role of so-called “upstream therapy.” To be of optimal benefit, such efforts should be undertaken before the onset of structural remodeling.34 Noninvasive detection of early signs of structural remodeling (e.g., delayed enhancement magnetic resonance imaging) may be of particular importance in identifying those patients at risk for developing further atrial fibrosis.
Conclusion Study limitations and clinical implications Limitations of the present study include small sample sizes, prior catheter ablations, use of autopsy controls, and use by some patients of drugs such as angiotensin-converting enzyme inhibitors, angiotensin receptor antagonists, and statins, which could decrease interstitial collagen deposition by blocking fibroproliferative signaling pathways.8 In addition, the numerically limited patient population could have negatively influenced some of the statistical results, particularly possible linear correlations between the various demographic, clinical, and morphologic parameters (e.g., patient age vs interstitial fibrosis). Furthermore, we cannot establish causal relationships between structural remodeling and arrhythmia, oxidative stress, or inflammation. As in any retrospective clinicopathologic study, we can only offer a snapshot of a given moment in the disease. Even with the limitations inherent in a retrospective analysis of patient tissue samples, our study clearly indicates that, despite the absence of underlying cardiovascular and pulmonary disease, patients with persistent LAF exhibit significant atrial structural remodeling and express molecular markers suggestive of atrial oxidative stress. The morphologic modifications are similar to but less severe than those affecting patients with mitral valve disease–related AF. The fact that all but 3 of the LAF patients had previously undergone 1 or more unsuccessful catheter ablations suggests that atrial structural remodeling had been widespread and of long duration.32,33 This hypothesis is supported by observations in a model of pacing-induced AF in goats in which atrial structural remodeling developed progressively over 16 weeks after atrial burst stimulation.20 In view of the
In patients with LAF, the atria are modified by structural remodeling and show molecular changes indicative of oxidative stress. Based on the literature9,20 and our present data, morphologic changes in LAF seem to occur early after the clinical onset of AF and are qualitatively no different than those observed in patients with AF associated to underlying cardiovascular diseases (e.g., mitral valve dysfunctions). These findings suggest that different types of AF are associated with the same spectrum of tissue lesions. Therefore, even patients labeled as having early LAF should undergo careful follow-up to discover early signs of atrial structural remodeling and, if possible, prevent the development of diffuse foci of interstitial fibrosis throughout the myocardial tissues.
Acknowledgments We thank Professor Rita Gatti, Department of Biomedical, Biotechnological, and Translational Sciences (S.Bi.Bi.T.), Unit of Histology, University of Parma, for expert confocal microscopy, and Ms. Maria Nicastro, Department of Clinical and Experimental Medicine, University of Parma, for technical assistance.
Appendix Supplementary data Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.hrthm. 2014.02.008.
References 1. Fuster V, Ryden LE, Cannom DS, et al. 2011 ACCF/AHA/HRS focused updates incorporated into the ACC/AHA/ESC 2006 guidelines for the management of patients with atrial fibrillation: a report of the American College of Cardiology
Corradi et al
2. 3.
4. 5.
6.
7.
8. 9.
10.
11.
12.
13.
14.
15. 16. 17.
Lone Atrial Fibrillation
Foundation/American Heart Association Task Force on practice guidelines. Circulation 2011;123:e269–e367. Kozlowski D, Budrejko S, Lip GY, et al. Lone atrial fibrillation: what do we know? Heart 2010;96:498–503. Camm AJ, Kirchhof P, Lip GY, et al. Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Europace 2010;12:1360–1420. Weijs B, Pisters R, Nieuwlaat R, et al. Idiopathic atrial fibrillation revisited in a large longitudinal clinical cohort. Europace 2012;14:184–190. Scardi S, Mazzone C, Pandullo C, Goldstein D, Poletti A, Humar F. Lone atrial fibrillation: prognostic differences between paroxysmal and chronic forms after 10 years of follow-up. Am Heart J 1999;137:686–691. Osranek M, Bursi F, Bailey KR, et al. Left atrial volume predicts cardiovascular events in patients originally diagnosed with lone atrial fibrillation: three-decade follow-up. Eur Heart J 2005;26:2556–2561. Daccarett M, McGann CJ, Akoum NW, MacLeod RS, Marrouche NF. MRI of the left atrium: predicting clinical outcomes in patients with atrial fibrillation. Expert Rev Cardiovasc Ther 2011;9:105–111. Corradi D, Callegari S, Maestri R, Benussi S, Alfieri O. Structural remodeling in atrial fibrillation. Nat Clin Pract Cardiovasc Med 2008;5:782–796. Ausma J, van der Velden HM, Lenders MH, et al. Reverse structural and gapjunctional remodeling after prolonged atrial fibrillation in the goat. Circulation 2003;107:2051–2058. Rienstra M, Hagens VE, Van Veldhuisen DJ, et al. Clinical characteristics of persistent lone atrial fibrillation in the RACE study. Am J Cardiol 2004;94: 1486–1490. Corradi D, Callegari S, Benussi S, et al. Regional left atrial interstitial remodeling in patients with chronic atrial fibrillation undergoing mitral-valve surgery. Virchows Arch 2004;445:498–505. Corradi D, Callegari S, Benussi S, et al. Myocyte changes and their left atrial distribution in patients with chronic atrial fibrillation related to mitral valve disease. Hum Pathol 2005;36:1080–1089. Corradi D, Callegari S, Maestri R, et al. Heme oxygenase-1 expression in the left atrial myocardium of patients with chronic atrial fibrillation related to mitral valve disease: its regional relationship with structural remodeling. Hum Pathol 2008;39:1162–1171. Corradi D, Callegari S, Maestri R, et al. Differential structural remodeling of the left-atrial posterior wall in patients affected by mitral regurgitation with or without persistent atrial fibrillation: a morphological and molecular study. J Cardiovasc Electrophysiol 2012;23:271–279. Lang RM, Bierig M, Devereux RB, et al. Recommendations for chamber quantification. Eur J Echocardiogr 2006;7:79–108. Corradi D, Maestri R, Macchi E, Callegari S. The atria: from morphology to function. J Cardiovasc Electrophysiol 2011;22:223–235. Frost L. Lone atrial fibrillation: good, bad, or ugly? Circulation 2007;115: 3040–3041.
9 18. Wyse DG. Idiopathic atrial fibrillation: a rose by any other name? Europace 2012;14:151–152. 19. White CW, Kerber RE, Weiss HR, Marcus ML. The effects of atrial fibrillation on atrial pressure-volume and flow relationships. Circ Res 1982;51:205–215. 20. Ausma J, Litjens N, Lenders MH, et al. Time course of atrial fibrillation-induced cellular structural remodeling in atria of the goat. J Mol Cell Cardiol 2001;33: 2083–2094. 21. Frustaci A, Chimenti C, Bellocci F, Morgante E, Russo MA, Maseri A. Histological substrate of atrial biopsies in patients with lone atrial fibrillation. Circulation 1997;96:1180–1184. 22. Kostin S, Klein G, Szalay Z, Hein S, Bauer EP, Schaper J. Structural correlate of atrial fibrillation in human patients. Cardiovasc Res 2002;54:361–379. 23. Hsieh MH, Lin YJ, Wang HH, et al. Functional characterization of atrial electrograms in a pacing-induced heart failure model of atrial fibrillation: importance of regional atrial connexin40 remodeling. J Cardiovasc Electrophysiol 2013;24:573–582. 24. Tsikas D. Analytical methods for 3-nitrotyrosine quantification in biological samples: the unique role of tandem mass spectrometry. Amino Acids 2012;42: 45–63. 25. Idriss NK, Blann AD, Lip GY. Hemoxygenase-1 in cardiovascular disease. J Am Coll Cardiol 2008;52:971–978. 26. Caglieri A, Goldoni M, De Palma G, et al. Exposure to low levels of hexavalent chromium: target doses and comparative effects on two human pulmonary cell lines. Acta Biomed 2008;79(Suppl 1):104–115. 27. Han W, Fu S, Wei N, et al. Nitric oxide overproduction derived from inducible nitric oxide synthase increases cardiomyocyte apoptosis in human atrial fibrillation. Int J Cardiol 2008;130:165–173. 28. Mihm MJ, Yu F, Carnes CA, et al. Impaired myofibrillar energetics and oxidative injury during human atrial fibrillation. Circulation 2001;104:174–180. 29. Guo Y, Lip GY, Apostolakis S. Inflammation in atrial fibrillation. J Am Coll Cardiol 2012;60:2263–2270. 30. Marcus GM, Smith LM, Ordovas K, et al. Intracardiac and extracardiac markers of inflammation during atrial fibrillation. Heart Rhythm 2010;7:149–154. 31. Liuba I, Ahlmroth H, Jonasson L, et al. Source of inflammatory markers in patients with atrial fibrillation. Europace 2008;10:848–853. 32. Corradi D, Callegari S, Gelsomino S, Lorusso R, Macchi E. Morphology and pathophysiology of target anatomical sites for ablation procedures in patients with atrial fibrillation: part I: atrial structures (atrial myocardium and coronary sinus). Int J Cardiol 2013;168:1758–1768. 33. Corradi D, Callegari S, Gelsomino S, Lorusso R, Macchi E. Morphology and pathophysiology of target anatomical sites for ablation procedures in patients with atrial fibrillation: part II: pulmonary veins, caval veins, ganglionated plexi, and ligament of Marshall. Int J Cardiol 2013;168:1769–1778. 34. Anne W, Duytschaever M. Upstream therapy in atrial fibrillation: traveling up the river to find the source. Heart Rhythm 2010;7:529–530.
Introduction “Lone” atrial fibrillation (LAF), also known as “idiopathic” atrial fibrillation (AF), is an ill-defined clinical entity Drs. Saffitz and Alfieri are co-senior authors. Address reprint requests and correspondence: Dr. Domenico Corradi, Department of Biomedical, Biotechnological, and Translational Sciences (S.Bi.Bi.T.), Unit of Pathology, University of Parma, Via Gramsci 14, 43126 Parma, Italy. E-mail address: [email protected].
1547-5271/$-see front matter B 2014 Heart Rhythm Society. All rights reserved.
remodeling and clinical parameters, including persistent LAF duration. CONCLUSION In persistent LAF patients, the atria are modified by structural remodeling and molecular changes of oxidative stress. Tissue changes in persistent LAF appear to occur early after its onset and are qualitatively no different than those observed in patients with atrial fibrillation related to conventional risk factors. These findings suggest that different types of atrial fibrillation are associated with the same spectrum of tissue lesions. Early intervention to restore sinus rhythm in persistent LAF patients may prevent irreversible tissue change, especially interstitial fibrosis. KEYWORDS Lone atrial fibrillation; Remodeling; Structure; Atrial cardiomyocyte; Pathology ABBREVIATIONS AF ¼ atrial fibrillation; Cx43 ¼ connexin43; HO-1 ¼ heme oxygenase-1; IL ¼ interleukin; LAF ¼ lone atrial fibrillation; NYHA ¼ New York Heart Association; PAS ¼ periodic acid Schiff; SOD-2 ¼ superoxide dismutase-2; TBARS ¼ thiobarbituric acid reactive substance; TNF-α ¼ tumor necrosis factor-α (Heart Rhythm 2014;0:1–9) rights reserved.
I
2014 Heart Rhythm Society. All
characterized mainly by the absence of any apparent explanation or underlying etiology, especially associated cardiopulmonary disorders including hypertension.1 Over the last few years, the number of patients with true LAF has decreased as new epidemiologic associations with AF have emerged (e.g., sleep apnea syndrome, obesity, high coffee or alcohol consumption, endurance sports activities, gene mutations).2 The reported prevalence of LAF has ranged widely between 2% and 30%, presumably reflecting the ambiguous
http://dx.doi.org/10.1016/j.hrthm.2014.02.008
2
Heart Rhythm, Vol 0, No 0, Month 2014
definition and uncertain diagnostic criteria for this supraventricular arrhythmia.3 However, new and more accurate clinical tests, as well as the discovery of new etiologies, have progressively reduced the number of LAF cases with no known cause.4 Patients with LAF do not represent a homogeneous group in terms of cardiovascular and thromboembolic risks. Instead, these patients should be divided into low-risk and high-risk categories, based on the chronicity of their arrhythmia and/or left atrial size, and treated accordingly.5,6 In this regard, an emerging parameter for a more accurate and individualized prognostic definition of LAF involves noninvasive assessment of atrial myocardial structural remodeling.7 Of the various histopathologic features of structural remodeling in AF substrates, myocardial fibrosis seems to be one of the most important, especially in view of its apparent structural stability and negligible regression after restoration of sinus rhythm.8,9 An interesting subset of LAF patients consists of cases that have progressed from paroxysmal to persistent AF in the absence of an underlying associated cardiopulmonary disease.10 To the best of our knowledge, no clinicopathologic studies on such patients have been reported in the English language literature. A thorough understanding of the structural and molecular state characterizing the LAF atrial myocardium certainly could represent the basis for future investigations focusing on structural remodeling prevention in these individuals. Based on recent advances regarding morphologic changes in AF and our own experience in analyzing the histopathology of AF atrial samples, the purpose of this retrospective investigation was to characterize the histopathology of tissue remodeling, quantify expression of molecular markers of oxidative stress and inflammation in left and right atrial endomyocardial biopsies, and correlate these changes with clinical characteristics in patients with persistent LAF undergoing surgical radiofrequency ablation.8,11–14
biopsies were obtained before any new radiofrequency ablation lines were produced through the atrial myocardium.12 Control autopsy tissues were sampled in the same way as those obtained surgically. Each left and right atrial sample was immediately divided into 2 fragments: 1 was fixed in 10% buffered formalin for 24 hours for histopathologic evaluations, and the other was frozen in liquid nitrogen and stored at –801C for subsequent molecular analyses.
Methods
Patient population and controls
Patient population and controls We studied 19 consecutive cases of LAF whose current arrhythmic episode lasted 47 days (“persistent AF” according to the current AF classification). Among these cases, patients whose present AF occurrence lasted Z1 year were further classified as having “long-lasting persistent AF.”1,3 As controls, we used 15 consecutive autopsied hearts from age-matched patients. The study was approved by the San Raffaele Hospital Institutional Review Board. All study patients gave written informed consent.
Atrial myocardial samples In each patient, myocardial samples (about 5 5 2 mm) were excised from the left atrial posterior wall (adjacent to the atriotomy line) and the right atrial free wall (between the 2 caval vein outlets).12 All of the biopsy sites were far from any previous transcatheter ablation lines. Furthermore,
Other methods Methods for histologic, morphometric, and immunofluorescence studies, enzyme-linked immunosorbent assay, and measurements of thiobarbituric acid reactive substance (TBARS), 3-nitrotyrosine levels, and heme oxygenase-1 (HO-1) mRNA are detailed in the Online Supplementary Data.
Statistical analysis Normality of data was assessed by the Shapiro-Wilks test. If there was no evidence against normality, data are expressed as mean ⫾ SD. Comparisons were made using independent (AF patients vs controls) or repeated measures (left atrium vs right atrium comparisons within each single patient or control case) Student t tests. When there were significant deviations from normality, data are expressed as median (with interquartile range), and nonparametric tests were performed (Mann-Whitney U test for independent samples and Wilcoxon test for repeated measures). The Pearson correlation test was applied when no outliers were present. Otherwise, the Spearman correlation test was used. Differences in connexin43 (Cx43) immunoreactive signal distribution were tested using the χ2 test. P o.05 was considered significant. Statistical analysis was performed using SPSS (version 21, IBM, Armonk, NY).
Results Nineteen patients (7.6% of the total AF patients undergoing surgical ablation at San Raffaele Hospital in the same period) fulfilled criteria for the diagnosis of persistent LAF and were enrolled in this study. According to their current AF episode duration, 11 patients (58%) had long-lasting persistent AF.3 General clinical data and preoperative echocardiographic measurements of the LAF patients are listed in Table 1. The average age of the LAF patients was 49 years. All but two were younger than 60 years, consistent with previous studies showing younger age of subjects with LAF than those with AF related to other conditions. There was a striking male predominance (17/19 cases [90%]), as also previously reported for LAF.4 Echocardiographic measures (Table 1) all were within normal limits except for left and right atrial volumes, which were above the reference values.15 Preoperatively, 5 patients had asymptomatic AF (New York Heart Association [NYHA] class I), whereas 14 had AF with palpitations (and on this basis only were considered as being
Corradi et al Table 1
Lone Atrial Fibrillation
3
General patient characteristics and echocardiographic data
No. of cases Age (y) Gender (M/F) Body mass index (kg/m2) BSA (m2) Total heart weight (g) Left atrial volume/BSA (mL/m2)* Right atrial volume/BSA (mL/m2) Left ventricular end-diastolic diameter (mm) left ventricular end-systolic diameter (mm) Left ventricular end-diastolic volume (mL) Left ventricular end-systolic volume (mL) Left ventricular ejection fraction (%) Interventricular septum thickness (mm) Left ventricular posterior wall thickness (mm) Pulmonary artery systolic pressure (mm Hg) Total LAF disease duration (mo) Current LAF episode duration (mo) No. of previous transcatheter ablation procedures (0/1/2/3/4/5) New York Heart Association functional class (I/II/III/IV)
LAF
Controls
P value
19 49 ⫾ 9 (31–66) 17/2 27 ⫾ 3 (18–29) 2.00 ⫾ 0.22 (1.44–2.40)
15 48 ⫾ 9 (33–60) 12/3 26 ⫾ 3 (16–29) 1.95 ⫾ 0.12 (1.38–2.05) 341 ⫾ 40 (270–390)
NS NS NS NS
42 ⫾ 13 (24–77) 34 ⫾ 15 (22–67) 50 ⫾ 4 (39–5) 29 ⫾ 3 (24–35) 95 ⫾ 10 (62–110) 40 ⫾ 6 (25–52) 58 ⫾ 4 (53–68) 10 ⫾ 1 (9–11) 9 ⫾ 1 (8–10) 25 ⫾ 4 (20–30) 91 ⫾ 69 (24–240) 17 ⫾ 19 (2–84) 3/4/4/4/3/1 5/14/0/0
Data are expressed as frequency or mean ⫾ SD (range). BSA ¼ body surface area; LAF ¼ lone atrial fibrillation. * Left atrial volume calculated using biplane Simpson method.
in NYHA class II). All patients underwent the Maze III surgical ablation procedure (13 through minithoracotomy and 6 through sternotomy). Sixteen patients (84%) had previously undergone a variable number (from 1–5, Table 1) of unsuccessful (nonsurgical) catheter ablations. After undergoing surgical AF ablation, all but 1 patient reverted to sinus rhythm. The preoperative pharmacologic treatments are listed in Table 2. All of the patients ate a standard Mediterranean-type diet. In accordance with their NYHA functional classification, all subjects performed mildto-moderate physical activity. General data regarding the control (autopsy) population are also listed in Table 1. Causes of death were traumatic injuries in 12 and nontraumatic cerebral hemorrhage in 3. Prior clinical measurements of atrial volume were not available for the autopsied controls. However, careful evaluation by experienced cardiopathologists indicated that atrial sizes in all control subjects fell within the normal range as previously described.16 Table 2
Preoperative treatment n (%)
Acetylsalicylic acid Allopurinol Amiodarone Angiotensin-converting enzyme inhibitor Angiotensin receptor antagonist Beta-blocker Calcium channel blocker Gastric acid pump inhibitor Oral anticoagulant Statin Data are expressed as no. of patients (percentage).
1 (5) 1 (5) 1 (5) 1 (5) 7 (37) 8 (42) 2 (10) 2 (10) 14 (74) 1 (5)
General histopathology and morphometric evaluation Atrial myocardium from patients with LAF showed significant pathologic changes consisting of hypertrophy, myocytolytic degeneration and glycogen accumulation in cardiomyocytes, remodeling of cardiomyocyte bundle orientation, and accumulation of interstitial and perivascular fibrosis. Representative images are shown in Figure 1, and quantitative morphometric data are given in Table 3. Atrial cardiomyocytes were sometimes arranged in irregular bundles (Figure 1A) compared to controls (Figure 1B). Myocytolytic degeneration was characterized by perinuclear loss of sarcomeres appearing as a clear vacuole extending eccentrically into the adjacent sarcoplasm (Figure 1C). These vacuoles were inconstantly filled with glycogen (seen as reddish material in the periodic acid–Schiff [PAS]-stained histologic sections in Figure 1D). Other cardiomyocytes without apparent loss of sarcomeres also showed moderately increased punctate sarcoplasmic PAS positivity, presumably representing intermyofibrillar glycogen accumulation secondary to cell metabolic changes. Atrial myocardial architecture in AF samples was typically modified by pathologic fibrosis between muscle cells (interstitial fibrosis; Figure 1E) and surrounding the adventitia of small intramural blood vessels (perivascular fibrosis; Figure 1F). In some cases, the subendocardial layer was thickened as a result of mild fibroelastosis. The pattern of fibrosis observed in AF patients was typical of what has been reported in previous studies of tissue remodeling in AF and was distinctly different than the dense local fibrosis resulting from previous catheter ablation procedures. Quantitative analysis of structural remodeling demonstrated significant differences between tissue features from
4
Heart Rhythm, Vol 0, No 0, Month 2014
Figure 1 Representative examples of left atrial structural remodeling in lone atrial fibrillation (LAF). Atria in LAF (A vs controls in B) show changes in cardiomyocyte size and orientation. Some cardiomyocytes show myolysis with characteristic perinuclear loss of sarcomeres (arrowheads in C) and glycogen accumulation (arrowhead in D). Interstitial fibrosis (arrowheads in A and E) and perivascular fibrosis (arrowheads in F) reflect collagen deposition around cardiomyocytes and intramural coronary branches. Bar ¼ 200 μm in A, B, E, and F; bar ¼ 100 μm in C and D.
Table 3
Morphometry LAF
Controls
LA
RA
LA †
RA
2.3 ⫾ 2.2 (0.9–10.1) 0 0 Interstitial fibrosis (%) 4.6 ⫾ 3.9 (1.1–16.9) 2.8 ⫾ 3.3 (0–10.5)† 0.5 ⫾ 0.1 (0–2.2) 0.4 ⫾ 0.2 (0–1.4) Cardiomyocyte myocytolysis (%) 5.1 ⫾ 5.6 (0–20.0)† Cardiomyocyte transverse 16.6 ⫾ 2.7 (12.9–20.6)† 15.5 ⫾ 2.6 (10.9–20.1)† 10.9 ⫾ 0.5 (10.4–13.4) 10.3 ⫾ 0.8 (10.2–13.5) diameter (mm) *†
Data are expressed as average ⫾ SD (range). LA ¼ left atrium; LAF ¼ lone atrial fibrillation; RA ¼ right atrium. * P o.05 (LA vs RA, in patients). † P o.0001 (patients vs controls).
Corradi et al
Lone Atrial Fibrillation
AF patients vs controls for all 3 parameters of tissue remodeling (myocytolysis, hypertrophy and fibrosis; Table 3). Controls showed no fibrosis and virtually no myocytolytic degeneration. PAS-positive material in controls consisted of minute particles distributed homogeneously throughout the sarcoplasm. Changes in cell and tissue structure generally appeared to be more severe in the left vs the right atrium in patients with AF, although this difference was significantly different only for the amount of interstitial fibrosis (Table 3). Cardiomyocyte transverse diameters showed statistically significant correlation between left atrium and right atrium (Pearson test, r ¼ 0.61, P o.05). There were no correlations between any of the histopathologic parameters and the number of previous transcatheter ablations. There was also no correlation between the amount of atrial fibrosis vs patient’s age or AF duration.
Connexin43 pattern Strong immunoreactive signal for Cx43 was seen in 16 of 19 control atrial cardiomyocytes both at intercalated disk regions and along the lateral sarcolemma (Figure 2). In the
5 remaining 3 cases, the amount of signal appeared to be reduced. By contrast, Cx43 signal was diminished in 11 of 19 of the LAF cases (P o.05 compared with controls; Figure 2). Signal was abundant at intercalated disks in samples from 7 AF cases and lateralized in 1 case. Reduced Cx43 signal occurred equally in both atria in 8 of the 11 cases of LAF with reduced signal. Signal was reduced in the left atrium but normal in the right atrium in 1 case and was reduced in the right atrium but normal in the left atrium in 2 cases.
Molecular markers of oxidative stress and inflammation Quantitative measurements of tissue markers of oxidative stress and inflammation are given in Table 4. Expression of the oxidative stress-related protein HO-1 was significantly increased in the left and right atrium (10- and 37-fold, respectively) in LAF patients compared to controls. Increased immunoreactive signal for HO-1 was also readily apparent in atrial cardiomyocytes from LAF patients compared to controls (Figure 3), whereas other markers of oxidative stress or inflammation showed no apparent
Figure 2 Connexin43 (Cx43) distribution in representative cases of lone atrial fibrillation (LAF) and controls. Controls showed strong immunoreactive signal at cell–cell junctions and in some cases at lateral membranes. In both LAF samples, Cx43 signal is significantly depressed. Bar ¼ 40 μm.
6 Table 4
Heart Rhythm, Vol 0, No 0, Month 2014 Proteins, thiobarbituric acid reactive substance, and 3-nytrotyrosine molecular measurements LAF
Controls
LA HO-1 (μg/g proteins) SOD-1 (μg/g proteins) SOD-2 (μg/g proteins) IL-6 (ng/g proteins) IL-8 (ng/g proteins) IL-10 (ng/g proteins) TNF-α (ng/g proteins) TBARS (nmol/g proteins) 3-Nitrotyrosine (mmol/mol tyrosine)
RA *
8.3 (3.0-19.8) 4.6 ⫾ 2.3 0.6 (0.3-1.1)† 26.4 (14.3-41.2) 25.8 ⫾ 16.3† 1.7 (1.1-5.4)† 25.3 ⫾ 16.8 16.5 (6.9-26.1)† 62.9 (46.2-89.5)*
*
14.7 (2.6-24.7) 4.7 ⫾ 2.9 1.2 (0.5-4.6) 41.6 (20.4-75.5) 36.1 ⫾ 27.4 4.3 (2.9-9.2) 37.3 ⫾ 26.3† 19.3 (11.0-79.9) 73.8 (50.4-188.4)*
LA
RA
0.8 (0.6-4.0) 5.2 ⫾ 3.2 1.7 (0.4-3.1) 50.3 (29.0-75.4) 47.7 ⫾ 26.8 5.0 (3.2-8.6) 31.7 ⫾ 17.8 36.1 (23.0-46.3) 28.8 (18.2-39.6) 22.4
0.4 (0.2-1.3) 3.5 ⫾ 1.3 0.7 (0.4-1.1) 40.6 (28.8-66.3) 40.5 ⫾ 27.5 3.6 (2.4-9.5) 19.1 ⫾ 9.6 26.2 (18.4-38.4) (17.0-27.5)
* ¼ p o 0.001, † ¼ p o 0.05. Data are expressed as average ⫾ SD or median value (interquartile range). HO ¼ heme oxygenase-1; IL ¼ interleukin; LA ¼ left atrium; LAF ¼ lone atrial fibrillation; RA ¼ right atrium; SOD ¼ superoxide dismutase; TBARS ¼ thiobarbituric acid reactive substance; TNF-α ¼ tumor necrosis factor-α.
differences in immunostaining compared with controls (data not shown). HO-1 mRNA was detected in both atria of patients with LAF in roughly equal amounts (0.47 ⫾ 0.33 vs 0.48 ⫾ 0.27 arbitrary units for left and right atria, respectively; P ¼ NS). Increased levels of HO-1 protein were associated with significantly greater levels (2- to 3fold) of the oxidative stress marker 3-nitrotyrosine in LAF tissue compared with controls (Table 4). Otherwise, left atrial tissue from LAF patients contained lower levels of superoxide dismutase-2 (SOD-2), interleukin (IL)-8, IL-10, and TBARS than the corresponding left atrial control tissue. Right atrial tissue from LAF patients contained greater amounts of tumor necrosis factor-α (TNF-α) than the corresponding right atrial control tissue (Table 4). The magnitude of these differences was not nearly as great, however, as that seen for HO-1 and 3nitrotyrosine.
Discussion In the present study, we found the same type of structural and molecular remodeling in the atria of patients with LAF as previously described in patients with more conventional disease-associated AF, thus suggesting that the different types of AF may be associated with the same continuum of lesions. The definition of LAF is still based largely on the apparent absence of other disorders that might account for “secondary” AF.2 Therefore, the diagnosis depends on the accuracy of clinical evaluation. Advances in such evaluations and identification of additional risk factors associated with AF have progressively diminished the fraction of cases that initially would have been labeled as “lone,”2 such that some now regard the designation of lone/idiopathic AF as obsolete.17 Nevertheless, the effects of the arrhythmia itself on atrial remodeling may be greater in LAF than in AF
Figure 3 Heme oxygenase-1 (HO-1) immunoreactive signal (green) is located in the cardiomyocyte sarcoplasm, identified by red staining for α-sarcomeric actin (SA). HO-1 signal intensity is much greater in the pathologic atrium than in the control (C). Yellow color indicates green and red signal superimposition. Bar ¼ 40 μm.
Corradi et al
Lone Atrial Fibrillation
associated with conventional risk factors (e.g., mitral valve disease) in which structural and molecular remodeling likely precede onset of AF. In this regard, pacing-induced AF in animals may be a better model of LAF than AF associated with usual risk factors.8 Age 460 years and increased atrial volume have generally been regarded as exclusion criteria for LAF, but recent studies have identified cases of AF without any evident underlying process in patients older than 60 years, suggesting that this age cutoff may be inappropriate.4,18 In the present study, we did not exclude patients with atrial enlargement in the absence of other indications of cardiovascular disease. In such cases, atrial enlargement may be the result rather than the cause of AF,10 although at some point in the natural history atrial enlargement likely is both a cause and a consequence of the arrhythmia. Structurally, the atrial myocardial architecture of our patient population had been altered by changes in cardiomyocytes and the interstitium, including increased cardiomyocyte size, myolysis, and interstitial fibrosis. All of these modifications were significantly greater in the LAF patients than in the control population but were less severe than those seen in our previous studies of AF associated with mitral valve disease. For example, the proportion of left and right atrial cardiomyocytes showing myolytic degeneration was 5.1% and 4.6%, respectively, in LAF compared with 15.1% to 20.8% and 7.16% to 10.4%, respectively, in non-lone forms of AF. Hypertrophy of atrial cardiomyocytes in LAF was only slightly less than that seen in valvular AF (16.6 vs 18.6–19 μm, respectively).11–14 The left atrium showed more interstitial fibrosis than the right atrium in LAF (4.6 ⫾ 3.9 vs 2.3 ⫾ 2.2, respectively, P o.05), perhaps related to the different hemodynamic loads of these 2 atrial chambers in AF.19 On the whole, these data confirm the fact that changes in AF structural remodeling are essentially nonspecific and suggest that, in terms of myocardial architectural distortion, the various forms of AF should be placed in a spectrum of disease where the specific effects of the arrhythmia (including its duration), associated diseases, and aging should be considered in addition to other factors. We found a consistent lack of correlation between AF duration and structural remodeling parameters in our previous studies on valvular AF (even after combining the different cases in a single patient population, unpublished data).11–14 The large variability in structural remodeling as a manifestation of myocardial injury likely is related to many factors, including genetics, therapy, and aging in addition to the usual risk factors implicated in AF initiation and/or perpetuation). It is unlikely that structural remodeling of atrial myocardium develops at the same rate over the natural history of AF. In fact, it has been demonstrated experimentally that histopathologic changes in develop progressively over an interval of 4 months of pacing-induced AF, by which time structural remodeling is fully established.20 Thus, it appears that the atria undergo structural remodeling early after the onset of AF, which might explain the consistent lack
7 of correlation between the severity of structural remodeling and the duration of AF in patients whose arrhythmia has occurred for at least a few months. Lastly, AF duration usually is calculated from the first arrhythmic episode, but the number of episodes over the clinical history of AF may be highly variable and often is unknown. The actual disease interval may vary considerably in patients with the same apparent AF duration. In contrast to the study by Frustaci et al,21 we did not find inflammatory infiltrates in myocardial samples from LAF patients. However, the patient populations were not identical, with paroxysmal AF in the previous investigation and persistent AF in the present study. We cannot definitively rule out the possibility that the fibrosis we observed could have been caused by previous myocarditis, and resulting substrate could have played a role in inducing and/or perpetuating the arrhythmia. Changes in the amount of immunoreactive signal for connexins and redistribution of signal from intercalated disk regions to the lateral margins of atrial cardiomyocytes have been described in AF.22,23 In our study, the amount of Cx43 signal was reduced in significantly more AF samples than controls, providing at least some basis for speculating that reduced coupling may have contributed to AF initiation and/ or maintenance. Because the number of LAF cases was limited, it was not possible to correlate changes in Cx43 signal with AF duration, left atrial dimensions, or the other features of structural and molecular remodeling analyzed in this study. Oxidative stress, in which reactive oxygen species damage both structural and functional proteins in cardiomyocytes, has been implicated in the pathogenesis of AF.8 In this study, we observed significantly increased expression of the protective enzyme HO-1 in LAF subjects compared to controls as well as higher levels of the oxidative stress marker 3-nitrotyrosine, which is formed when reactive nitrogen species (e.g., peroxynitrite, nitryl chloride) interact with the aromatic ring of tyrosine in soluble amino acids and in proteins.24 Previously, we reported increased HO-1 expression (greater in the left atrial posterior wall that in the left atrial appendage) in 2 different patient populations with valvular AF.13,14 HO-1 expression has previously been shown to confer protection in various models of cardiovascular injury or disease, including hypertension, atherosclerosis, myocardial infarction, and allograft rejection.25 Further and more focused investigations are necessary to test whether HO-1 may serve as a predictive marker in AF. In the present study, HO-1 expression was not associated with a parallel increase in SOD-2 or lipid peroxidation (as determined by measurement of TBARS) in the left atrial myocardium. In fact, TBARS levels were significantly decreased compared to controls. This might reflect the activation of antioxidative stress-protective mechanisms, as observed in in vitro models.26 The significantly increased levels of 3-nitrotyrosine also suggest greater atrial nitric oxide levels and ongoing nitric oxide–dependent reactive species-induced oxidative stress.27 Increased myocardial
8 3-nitrotyrosine has previously been reported in AF associated with mitral valve disease.28 Inflammatory markers such as C-reactive protein, TNF-α, IL-2, IL-6, and IL-8 are thought to play a role in initiating and perpetuating AF and AF-related thrombosis through mechanisms involving endothelial damage/dysfunction and platelet activation.29 Whether AF is the cause or a consequence of inflammation is not known, but it is likely that both relationships occur. Inflammation may be a potent trigger of AF, but at the same time AF seems to promote an inflammatory and prothrombotic state.8,29,30 Only limited data are available on the systemic or local nature of inflammation in AF. Liuba et al31 found higher plasma levels of IL-8 in femoral vein, right atrial, and coronary sinus blood, but not in pulmonary vein samples, consistent with a systemic source of inflammation. Marcus et al30 measured higher C-reactive protein levels in the left atrium than in the corresponding coronary sinus and suggested that transcardiac cytokine gradients in AF arise by sequestration of inflammatory cytokines in the heart. The results of the present study do not indicate increased local production/ accumulation of atrial myocardial IL-6, IL-8, IL-10, and TNF-α in LAF patients.
Heart Rhythm, Vol 0, No 0, Month 2014 high stability of collagen fibers, preventing or limiting interstitial fibrosis appears to be an important way to mitigate AF. In this regard, interstitial fibrosis was the only component of structural remodeling that had not regressed 4 months after restoration of sinus rhythm in the goal model of pacinginduced AF.9 Clinical and experimental evidence suggests that interstitial fibrosis and other components of atrial structural remodeling interact with AF in a vicious cycle in which they act as both a cause and a consequence of each other.8 In this view, progression of the abnormal substrate would play a part in the transition from paroxysmal to persistent AF. In view of limitations associated with antiarrhythmic drugs and extensive ablative procedure and with increasing evidence of ongoing oxidative stress, attention has turned to the role of so-called “upstream therapy.” To be of optimal benefit, such efforts should be undertaken before the onset of structural remodeling.34 Noninvasive detection of early signs of structural remodeling (e.g., delayed enhancement magnetic resonance imaging) may be of particular importance in identifying those patients at risk for developing further atrial fibrosis.
Conclusion Study limitations and clinical implications Limitations of the present study include small sample sizes, prior catheter ablations, use of autopsy controls, and use by some patients of drugs such as angiotensin-converting enzyme inhibitors, angiotensin receptor antagonists, and statins, which could decrease interstitial collagen deposition by blocking fibroproliferative signaling pathways.8 In addition, the numerically limited patient population could have negatively influenced some of the statistical results, particularly possible linear correlations between the various demographic, clinical, and morphologic parameters (e.g., patient age vs interstitial fibrosis). Furthermore, we cannot establish causal relationships between structural remodeling and arrhythmia, oxidative stress, or inflammation. As in any retrospective clinicopathologic study, we can only offer a snapshot of a given moment in the disease. Even with the limitations inherent in a retrospective analysis of patient tissue samples, our study clearly indicates that, despite the absence of underlying cardiovascular and pulmonary disease, patients with persistent LAF exhibit significant atrial structural remodeling and express molecular markers suggestive of atrial oxidative stress. The morphologic modifications are similar to but less severe than those affecting patients with mitral valve disease–related AF. The fact that all but 3 of the LAF patients had previously undergone 1 or more unsuccessful catheter ablations suggests that atrial structural remodeling had been widespread and of long duration.32,33 This hypothesis is supported by observations in a model of pacing-induced AF in goats in which atrial structural remodeling developed progressively over 16 weeks after atrial burst stimulation.20 In view of the
In patients with LAF, the atria are modified by structural remodeling and show molecular changes indicative of oxidative stress. Based on the literature9,20 and our present data, morphologic changes in LAF seem to occur early after the clinical onset of AF and are qualitatively no different than those observed in patients with AF associated to underlying cardiovascular diseases (e.g., mitral valve dysfunctions). These findings suggest that different types of AF are associated with the same spectrum of tissue lesions. Therefore, even patients labeled as having early LAF should undergo careful follow-up to discover early signs of atrial structural remodeling and, if possible, prevent the development of diffuse foci of interstitial fibrosis throughout the myocardial tissues.
Acknowledgments We thank Professor Rita Gatti, Department of Biomedical, Biotechnological, and Translational Sciences (S.Bi.Bi.T.), Unit of Histology, University of Parma, for expert confocal microscopy, and Ms. Maria Nicastro, Department of Clinical and Experimental Medicine, University of Parma, for technical assistance.
Appendix Supplementary data Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.hrthm. 2014.02.008.
References 1. Fuster V, Ryden LE, Cannom DS, et al. 2011 ACCF/AHA/HRS focused updates incorporated into the ACC/AHA/ESC 2006 guidelines for the management of patients with atrial fibrillation: a report of the American College of Cardiology
Corradi et al
2. 3.
4. 5.
6.
7.
8. 9.
10.
11.
12.
13.
14.
15. 16. 17.
Lone Atrial Fibrillation
Foundation/American Heart Association Task Force on practice guidelines. Circulation 2011;123:e269–e367. Kozlowski D, Budrejko S, Lip GY, et al. Lone atrial fibrillation: what do we know? Heart 2010;96:498–503. Camm AJ, Kirchhof P, Lip GY, et al. Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Europace 2010;12:1360–1420. Weijs B, Pisters R, Nieuwlaat R, et al. Idiopathic atrial fibrillation revisited in a large longitudinal clinical cohort. Europace 2012;14:184–190. Scardi S, Mazzone C, Pandullo C, Goldstein D, Poletti A, Humar F. Lone atrial fibrillation: prognostic differences between paroxysmal and chronic forms after 10 years of follow-up. Am Heart J 1999;137:686–691. Osranek M, Bursi F, Bailey KR, et al. Left atrial volume predicts cardiovascular events in patients originally diagnosed with lone atrial fibrillation: three-decade follow-up. Eur Heart J 2005;26:2556–2561. Daccarett M, McGann CJ, Akoum NW, MacLeod RS, Marrouche NF. MRI of the left atrium: predicting clinical outcomes in patients with atrial fibrillation. Expert Rev Cardiovasc Ther 2011;9:105–111. Corradi D, Callegari S, Maestri R, Benussi S, Alfieri O. Structural remodeling in atrial fibrillation. Nat Clin Pract Cardiovasc Med 2008;5:782–796. Ausma J, van der Velden HM, Lenders MH, et al. Reverse structural and gapjunctional remodeling after prolonged atrial fibrillation in the goat. Circulation 2003;107:2051–2058. Rienstra M, Hagens VE, Van Veldhuisen DJ, et al. Clinical characteristics of persistent lone atrial fibrillation in the RACE study. Am J Cardiol 2004;94: 1486–1490. Corradi D, Callegari S, Benussi S, et al. Regional left atrial interstitial remodeling in patients with chronic atrial fibrillation undergoing mitral-valve surgery. Virchows Arch 2004;445:498–505. Corradi D, Callegari S, Benussi S, et al. Myocyte changes and their left atrial distribution in patients with chronic atrial fibrillation related to mitral valve disease. Hum Pathol 2005;36:1080–1089. Corradi D, Callegari S, Maestri R, et al. Heme oxygenase-1 expression in the left atrial myocardium of patients with chronic atrial fibrillation related to mitral valve disease: its regional relationship with structural remodeling. Hum Pathol 2008;39:1162–1171. Corradi D, Callegari S, Maestri R, et al. Differential structural remodeling of the left-atrial posterior wall in patients affected by mitral regurgitation with or without persistent atrial fibrillation: a morphological and molecular study. J Cardiovasc Electrophysiol 2012;23:271–279. Lang RM, Bierig M, Devereux RB, et al. Recommendations for chamber quantification. Eur J Echocardiogr 2006;7:79–108. Corradi D, Maestri R, Macchi E, Callegari S. The atria: from morphology to function. J Cardiovasc Electrophysiol 2011;22:223–235. Frost L. Lone atrial fibrillation: good, bad, or ugly? Circulation 2007;115: 3040–3041.
9 18. Wyse DG. Idiopathic atrial fibrillation: a rose by any other name? Europace 2012;14:151–152. 19. White CW, Kerber RE, Weiss HR, Marcus ML. The effects of atrial fibrillation on atrial pressure-volume and flow relationships. Circ Res 1982;51:205–215. 20. Ausma J, Litjens N, Lenders MH, et al. Time course of atrial fibrillation-induced cellular structural remodeling in atria of the goat. J Mol Cell Cardiol 2001;33: 2083–2094. 21. Frustaci A, Chimenti C, Bellocci F, Morgante E, Russo MA, Maseri A. Histological substrate of atrial biopsies in patients with lone atrial fibrillation. Circulation 1997;96:1180–1184. 22. Kostin S, Klein G, Szalay Z, Hein S, Bauer EP, Schaper J. Structural correlate of atrial fibrillation in human patients. Cardiovasc Res 2002;54:361–379. 23. Hsieh MH, Lin YJ, Wang HH, et al. Functional characterization of atrial electrograms in a pacing-induced heart failure model of atrial fibrillation: importance of regional atrial connexin40 remodeling. J Cardiovasc Electrophysiol 2013;24:573–582. 24. Tsikas D. Analytical methods for 3-nitrotyrosine quantification in biological samples: the unique role of tandem mass spectrometry. Amino Acids 2012;42: 45–63. 25. Idriss NK, Blann AD, Lip GY. Hemoxygenase-1 in cardiovascular disease. J Am Coll Cardiol 2008;52:971–978. 26. Caglieri A, Goldoni M, De Palma G, et al. Exposure to low levels of hexavalent chromium: target doses and comparative effects on two human pulmonary cell lines. Acta Biomed 2008;79(Suppl 1):104–115. 27. Han W, Fu S, Wei N, et al. Nitric oxide overproduction derived from inducible nitric oxide synthase increases cardiomyocyte apoptosis in human atrial fibrillation. Int J Cardiol 2008;130:165–173. 28. Mihm MJ, Yu F, Carnes CA, et al. Impaired myofibrillar energetics and oxidative injury during human atrial fibrillation. Circulation 2001;104:174–180. 29. Guo Y, Lip GY, Apostolakis S. Inflammation in atrial fibrillation. J Am Coll Cardiol 2012;60:2263–2270. 30. Marcus GM, Smith LM, Ordovas K, et al. Intracardiac and extracardiac markers of inflammation during atrial fibrillation. Heart Rhythm 2010;7:149–154. 31. Liuba I, Ahlmroth H, Jonasson L, et al. Source of inflammatory markers in patients with atrial fibrillation. Europace 2008;10:848–853. 32. Corradi D, Callegari S, Gelsomino S, Lorusso R, Macchi E. Morphology and pathophysiology of target anatomical sites for ablation procedures in patients with atrial fibrillation: part I: atrial structures (atrial myocardium and coronary sinus). Int J Cardiol 2013;168:1758–1768. 33. Corradi D, Callegari S, Gelsomino S, Lorusso R, Macchi E. Morphology and pathophysiology of target anatomical sites for ablation procedures in patients with atrial fibrillation: part II: pulmonary veins, caval veins, ganglionated plexi, and ligament of Marshall. Int J Cardiol 2013;168:1769–1778. 34. Anne W, Duytschaever M. Upstream therapy in atrial fibrillation: traveling up the river to find the source. Heart Rhythm 2010;7:529–530.
