Histological Substrate of Atrial Biopsies in Patients With Lone Atrial Fibrillation
Background Lone atrial fibrillation (LAF) is a common clinical syndrome, but its origin remains unknown.
Methods and Results We performed endomyocardial biopsies of the right atrial septum (2 to 3 per patient; mean, 2.8) and of the two ventricles (6 per patient) in 12 patients (10 men, 2 women; mean age, 32 years) with paroxysmal LAF refractory to conventional antiarrhythmic treatment. As controls, we used endomyocardial biopsies (3 to 5 per patient; mean, 4.4) from the right atrial septum of 11 patients with Wolff-Parkinson-White syndrome (WPW) undergoing resection of the abnormal AV pathway. The weight of the biopsies ranged from 2.8 to 4.5 mg. Biopsy samples were processed for histology and electron microscopy and were read by a pathologist blinded to clinical data. All patients underwent two-dimensional Doppler echocardiography; cardiac catheterization; coronary angiography; and hormonal, virologic, and electrophysiological studies. All tests and WPW biopsies were normal, but all LAF atrial biopsy specimens (average, 2.8 per patient) showed abnormalities (P<.0001). The type of abnormalities varied: Two patients had a severe hypertrophy with vacuolar degeneration of the atrial myocytes and ultrastructural evidence of fibrillolysis occupying >50% of the areas assessed morphometrically (P=.50), 8 had lymphomononuclear infiltrates with necrosis of the adjacent myocytes (5 with fibrosis and 3 without; P<.003), and 2 had only nonspecific patchy fibrosis (P=.50). Biventricular biopsies were abnormal in only 3 patients and showed inflammatory infiltrates similar to those found in atrial biopsies.
Conclusions Abnormal atrial histology was uniformly found in multiple biopsy specimens in all patients with LAF. It was compatible with a diagnosis of myocarditis in 66% of patients (active in 25%) and of noninflammatory localized cardiomyopathy in 17% and was represented by patchy fibrosis in 17%. The cause of the pathological changes, which were found only in atrial septal biopsies but not in biventricular biopsies, in 75% of patients remains unknown.
Atrial fibrillation is the most common sustained arrhythmia in clinical practice. Although it is usually associated with recognizable organic heart disease or hyperthyroidism, it may occur without clinically evident abnormalities. This isolated form, commonly called LAF, appears in 2.7%1 to 11.4%2 of cases, depending on the age of the patients.
The cause of LAF is poorly understood, and to the best of our knowledge, there are no atrial biopsy or postmortem studies reported in the literature on this specific topic. Ventricular endomyocardial biopsy findings observed in patients with chronic LAF are mostly nonspecific,3 and it is not known to what extent they reflect the atrial myocardial substrate.
After obtaining approval from the Ethics Committee of our institution and written informed consent from the patients, we performed right atrial and biventricular endomyocardial biopsies on 12 patients with paroxysmal LAF refractory to conventional antiarrhythmic therapy to study the atrial histological substrate and its relation to biventricular histological findings as a possible indication for treatment.
The study group consisted of 12 patients (10 men and 2 women; mean age, 32 years; range, 19 to 43 years) referred to our institution for treatment of symptomatic, refractory paroxysmal LAF (ie, of <24-hour duration). The patients were enrolled in a prospective study, approved by the Ethics Committee of our institution, from October 1993 through December 1995. According to the inclusion criteria, all patients had more than one episode of AF per month (range, 3 to 90) and had been refractory to at least two prior empirical drug regimens (range. two to five), including propafenone, quinidine, dysopiramide, sotalol, and flecainide (see the Table⇓). Hypertension, smoking, and drug or alcohol abuse were not apparent from the clinical histories.
All patients underwent routine laboratory tests, including those for glycemia, creatinemia, blood urea nitrogen, serum electrolytes (Na+, K+, Ca2+, and Mg2+), transaminase (sGOT and sGPT), erythrocyte sedimentation rate, and blood cell count. Hormonal screening included the assessment of catecholamines, metanephrines, vanillyl-mandelic, homovanillic, and 5-OH-indolacetic acid content in the urine over 24 hours; thyroid function tests (T3, T4, and thyroid-stimulating hormone); and adrenal scans. Serological tests for cardiotropic viruses included a search for echo, coxsackievirus B, influenza, and parainfluenza.
Cardiac studies included both noninvasive tests (ie, ECG with Holter monitoring and two-dimensional echocardiography with Doppler analysis) and invasive procedures (ie, transesophageal electrophysiologic study, cardiac catheterization, left and right ventricular and coronary angiographies, and right atrial septal and biventricular endomyocardial biopsies). All patients underwent TAS while in sinus rhythm after administration of all antiarrhythmic drugs had been stopped for at least five half-lives. No patient was on long-term amiodarone therapy. TAS was performed by use of a moderately aggressive protocol: with single and double extrastimuli during sinus rhythm and during 600- and 400-ms cycle length pacing and 8-second atrial bursts at rates from 180 to 300 bpm. Patients with inducible, sustained (>1-minute duration) PAF at the time of basal TAS underwent serial electropharmacological testing with oral propafenone, sotalol, and flecainide to identify responders (ie, patients in whom sustained PAF was no longer inducible) and nonresponders (ie, those in whom it was still inducible). Responders were discharged on a drug regimen that prevented induction of PAF. Nonresponders or those in whom PAF was not inducible at the time of basal TAS were discharged on empirical amiodarone therapy. Patients with a histological pattern of active myocarditis were discharged on steroid therapy. All patients were followed up at 4-week intervals. At each visit, they were questioned regarding the efficacy and toxicity of drugs and underwent both physical examination and Holter monitoring. In patients with paroxysmal symptoms whose ECG and Holter recordings showed no evidence of recurrent PAF, transtelephonic recordings were obtained to document the correlation of symptoms and recurrence of arrhythmias.
Endomyocardial biopsy was performed in the septal region of the right atrium and in the septal-apical segment of both ventricles. In particular, atrial specimens were drawn from areas adjacent to the fossa ovalis, which was approached by a 7F (501-613A Cordis) long sheet and identified on a biplane (frontal and laterolateral projection) radiograph view with flashing of contrast medium. Two or three fragments from the atrium and three samples from each ventricular chamber were taken from each patient and processed for histology and TEM. Controls were biopsies taken from the right atrial septum of 11 patients with Wolff-Parkinson-White syndrome and no history of AF at the time of surgical resection of the abnormal AV pathway. The weight of the biopsies ranged from 2.8 to 4.5 mg. The pathologist was blinded to the cardiac site of the biopsies and the clinical history of LAF patients.
For TEM, all samples of myocardial tissue were fixed in a solution of 2% glutaraldehyde in a 0.1-mol/L phosphate buffer, pH 7.3, and embedded in an Epon resin following a standard schedule. Ultrathin sections were stained with uranyl acetate and lead citrate. The tissue for light microscopy was fixed in 10% buffered formalin and embedded in paraffin wax. Sections (5 μm thick) were cut and stained with hematoxylin and eosin, Miller’s elastic van Gieson’s,4 and Masson’s trichrome stain.
Morphometry included evaluation of both the diameter of myocardial fibers and the percentage of the myofibrillar component in the cytoplasmic area. The diameter of the myocardial fibers was measured either at the level of the nuclear region on micrographs at a magnification of ×460 or directly by a wide-field measuring optical microscope (Carl Zeiss Co). Morphometric analysis of the myofibrillar areas was carried out with a computer-assisted image analyzer, with KS-300 software (Carl Zeiss Co) directly on photographic negatives of TEM sections. The procedure occasionally required the operator’s assistance to ensure the delimitation of the boundaries between the areas of cytoplasm with and without myofibrils. When this problem was observed, demarcation of such areas was established with an electronic pencil. To reduce errors, we used at least two independent operators for the procedure. Delimitation of the boundaries is crucial to calculate the average cytoplasmic area belonging to myofibrillar material. Measurements were performed on a total of 20 micrographs (×4900), either photographic negatives or printed positives, taken from three different serial sections of each TEM sample; in each biopsy, a total of 50 myocardial cells were studied.
Discontinuous variables between patients and control subjects were compared by the Yates-corrected χ2 method. Continuous variables were compared by Student’s t test. A value of P<.05 was considered significant.
Routine blood and chemical analyses were within normal limits in all patients. In particular, thyroid function tests, catecholamines, and vanillyl-mandelic acid content in the urine over 24 hours were in the normal range. Serological tests for cardiotropic viruses were negative.
ECG recordings and 24-hour Holter monitoring showed several episodes of PAF in each patient (up to 20 per 24 hours). No conduction abnormalities, alterations of the ST segment or T wave, or ventricular arrhythmias were recorded.
In all cases, two-dimensional echocardiograms showed normal atrial and ventricular dimensions (left atrial diameter <40 mm, left ventricular end-diastolic diameter <56 mm, and right ventricular end-diastolic diameter <22 mm) and normal contractile function (left ventricular ejection fraction >55%) in all cases. AV valvular motion was also normal.
In all patients, invasive cardiac studies showed normal pulmonary and intracavity pressures (left ventricular end-diastolic pressure <12 mm Hg and right ventricular end-diastolic pressure <8 mm Hg), normal left and right ventriculograms, and normal coronary angiograms. Neither cardiac catheterization or biopsy produced any complications. One patient developed a PAF during the study, but it resolved spontaneously a few hours later.
Histological and Ultrastructural Findings
Optical and electron microscopic findings were normal in all 45 biopsy specimens from Wolff-Parkinson-White syndrome patients and abnormal in all 34 atrial biopsies from LAF patients (P<.0001), whereas only 4 of 72 ventricular specimens showed abnormal histology. The type of atrial abnormality was uniform in all 3 biopsies in nine patients—in 2 of 3 biopsies in one patient and in both specimens in the remaining two patients—but varied in the population.
In eight patients, inflammatory lymphomononuclear infiltrates, associated with focal necrosis of adjacent myocytes, were observed (P<.003 patients versus control subjects). According to the Dallas criteria,5 these findings are compatible with a diagnosis of atrial myocarditis. In three patients, the inflammatory changes were not associated with fibrosis (ie, they are compatible with active atrial myocarditis; Fig 1⇓); they were documented only in atrial specimens, and the LAF was of recent onset (the Table⇑). Five patients also had interstitial fibrosis with focal replacement (Fig 2⇓): in three, similar focal inflammatory lesions with interstitial fibrosis were also observed in ventricular specimens (see the Table⇑).
In two patients, striking areas of hypertrophy (mean cell diameter, 28.7 and 25.3 μm compared with a mean of 8.8 μm in control samples) were observed, with vacuolar degeneration of atrial myocytes associated with interstitial and focal replacement fibrosis without inflammatory infiltrates (Fig 3⇓). Such alterations were confined to the three atrial samples and were absent in all six biventricular specimens (Figs 4⇓ and 5⇓). TEM showed that atrial myocyte vacuoles were associated with areas in which myofibrillar material was either absent or fragmented, the space being occupied by amorphous or granular material with a few membranaceous organules. Morphometry showed a decrease of up to 60% in the myofibrillar area compared with control atrial myocardiocytes with no apparent quantitative changes in other components. The myofibrillar area (ie, the percentage of area related to myofibrils in the total area of a myocardiocyte) appeared smaller (14.3±3.4%) compared with ventricular myocardiocytes in the same patient (31.8±5.5%) and atrial control specimens (41.4±3.2%). Statistical evaluation of the morphometric values was highly significant (P<.0001 for patient atrial compared with ventricular myocardiocytes; P<.0003 for patient atrial myocardiocytes versus control specimens). The residual myofibrillar area appeared to be organized normally, with moderately contracted sarcomeres and no contraction bands or other abnormal features. Some swelling with fragmentation of cristae was apparent in the mitochondria, whereas the remaining organules had no degenerative changes. Ventricular biopsy showed no ultrastructural or morphometric abnormalities. These findings are compatible with a cardiomyopathic process confined to the atrial myocardium.
Finally, extensive areas of atrial fibrosis without inflammatory infiltrates were observed in the last two cases despite normal biventricular biopsies.
Electrophysiologic Study, Treatment, and Follow-up
The main clinical and electrophysiological features of the patients are summarized in the Table⇑. Three patients in whom PAF was not inducible during basal TAS were discharged on chronic amiodarone therapy. PAF was inducible in nine patients during basal TAS: four responders were discharged on chronic antiarrhythmic therapy (two on propafenone, one on sotalol, and one on flecainide); two nonresponders were discharged on chronic amiodarone therapy; and the three nonresponders with a histological pattern of active myocarditis were discharged on chronic steroid therapy (prednisone 1 mg · kg−1 · d−1 for 4 weeks tapered to 0.33 mg for 4 months).
During follow-up (mean, 12 months; range, 8 to 24 months), AF recurred in all nine patients on chronic antiarrhythmic therapy: five on amiodarone (three with noninducible PAF and two nonresponders), two on propafenone, one on flecainide, and one on sotalol. In patients with active myocarditis treated with steroids, AF did not recur. No significant changes in either ventricular dimension or contractility were documented.
In the two patients with cardiomyopathic changes and in one patient with myocarditis and fibrosis, moderate dilatation of the left atrium was observed (45, 48, and 44 mm in patients 1, 6, and 12, respectively; the Table⇑).
To the best of our knowledge, this is the first study to show that LAF is consistently associated with atrial histological abnormalities.
The histological findings in biopsies taken from the interatrial septum were compatible with a diagnosis of myocarditis in 66% of patients, with a noninflammatory cardiomiopathic process in 17%, and with patchy fibrosis in the remaining 17%. The histological findings of biventricular biopsies were abnormal only in 25% of patients and confirmed the finding of atrial myocarditis.
In eight patients, inflammatory infiltrates meeting the Dallas criteria for myocarditis were observed in atrial biopsies: three without fibrosis (active atrial myocarditis) and five associated with focal replacement fibrosis. Inflammatory lymphomononuclear infiltrates suggest a viral or postviral autoimmune mechanism. In only three patients were such infiltrates also observed in ventricular biopsies. Serological tests were negative for the most common viruses affecting the heart; however, in an extensive virological study of patients with clinical and histological evidence of myocarditis, a positive serological test was found in 50% of patients.6 Interstitial inflammatory infiltrates previously observed in the atrial myocardium of patients with chronic AF cannot be considered diagnostic of myocarditis because they were not associated with the presence of necrotic myocardiocytes surrounded by clusters of lymphocytes adherent to the sarcolemmal membrane as typically abserved in our patients (Figs 1⇑ and 2⇑). Only this association meets the Dallas criteria for myocarditis. Thus, it cannot be considered a consequence of the arrhythmia itself and suggests that the inflammatory process was a likely cause of the arrhythmia. Indeed, in the three patients with active atrial myocarditis, steroid treatment was associated with the absence of a recurrence of LAF, which occurred in all the patients on electrophysiological study (EPS)–guided antiarrhythmic therapy.
In two patients, a noninflammatory cardiomyopathy, with areas of striking hypertrophy and vacuolar degeneration of atrial myocytes but with normal ventricular histology compatible with a cardiomyopathic process confined to the atrial myocardium, was observed. It can be argued that these changes may be the result rather than the cause of AF; however, we found no relationship between PAF duration and severity of atrial degenerative changes; furthermore, the ultrastructural findings obtained in experimental studies inducing AF in animals exhibited different abnormalities, consisting of increases in mitochondrial size and number and in the disruption of sarcoplasmic reticulum.7 A cardiomyopathic process confined to the atrial myocytes may seem unusual, but other pathological processes of limited extension are known to occur in both the atrial (lipomatosis of interatrial septum)8 and ventricular (right ventricular cardiomyopathy)9 10 myocardium. Furthermore, the possibility of an isolated atrial cardiomyopathy has already been raised in a previous biopsy report11 that stressed the discrepancy between atrial and ventricular histological findings in some patients with idiopathic supraventricular arrhythmias.
Finally, the patchy areas of fibrous replacement (not observed in control biopsies) documented in the last two patients are nonspecific and may be the result of myocardial healing caused by different mechanisms (ie, toxic or inflammatory).
Thus, most often the histological substrate is strictly localized in the atrial myocardium. During follow-up, an isolated increase in left atrial dimension was observed in the two patients with atrial cardiomyopathy and in one of the eight patients with myocarditis, but ventricular dimension and contractile function remained normal in all patients. The cause of the atrial histological abnormality remains unknown because all tests, including serology for the most common viruses affecting the heart, were negative. However, each of the histological abnormalities observed (atrial myocyte degeneration and necrosis and patchy fibrosis) may represent an organic substrate for the electrogenic mechanisms involved in paroxysmal LAF. A previous, large postmortem study on patients with AF associated with organic heart disease12 showed a diversity of histological abnormalities involving diffusely the right and left atria. Although no postmortem study of patients with paroxysmal LAF is available, it is likely that in such isolated presentations of AF, the histological abnormality may diffusely involve the right and left atria. Our data seem consistent with this possibility. Obviously, no biopsy study can provide the kind of information on the extension of a myocardial disease that could be obtained at postmortem. However, a previous comparison of biopsy and autopsy material13 showed adequate representation of the entire interventricular septum and, though less consistently, of the homologous ventricular free wall by five biopsies from the interventricular septum. The high rate of abnormal atrial specimens (100%) and the uniformity of the lesion in each patient (97%) compared with the low rate of abnormal ventricular specimens (5%) and lack of abnormal atrial controls suggest that the abnormalities may be diffusely present in the atrial tissue of patients with LAF. Thus, coherent results of biopsies taken from the right atrial septum may also be representative of the atrial myocardium.
The diagnostic contribution of biventricular biopsy with three specimens per chamber is of limited value because abnormal findings similar to those observed in atrial biopsies were found in only 3 of the 12 patients studied. The low diagnostic value of ventricular endomyocardial biopsies3 is not surprising because combined atrial and ventricular biopsies for various atrial bradyarrhythmias and tachyarrhythmias11 14 15 have shown a poor correlation between atrial and ventricular findings. In the past, atrial biopsy for lone atrial arrhythmias has been limited to single cases14 15 and abstract11 reports because of the risk of perforating the thin structure of the atrial wall. However, the established routine of the right atrial septal approach for left atrial catheterization and mitral valvuloplasty suggests that this anatomic region can be considered reasonably safe for obtaining atrial biopsy specimens.
The results of atrial septal biopsies may help predict a favorable outcome when consistent with active myocarditis. In such patients, steroids were used on the basis of the positive results obtained with steroids and immunosuppression in ventricular arrhythmias associated with myocarditis.16 17 However, it is not possible to tell whether the absence of recurrences was related to the therapy or to the spontaneous healing of the acute disease process. Conversely, TAS appears inadequate for predicting the response to be expected from chronic antiarrhythmic therapy and confirms the low efficacy of anthiarrhythmic drugs in preventing a recurrence of PAF.
In conclusion, our study provides evidence of histological abnormalities as a consistent organic substrate for paroxysmal LAF refractory to conventional therapy. In this patient population, we perform atrial biopsies, on an institutional basis, to obtain more information on the cause of and identification of patients with active atrial myocarditis that is potentially susceptible to specific medical treatment.
Selected Abbreviations and Acronyms
|LAF||=||lone atrial fibrillation|
|PAF||=||paroxysmal atrial fibrillation|
|TAS||=||transesophageal electrophysiological study|
|TEM||=||transmission electron microscopy|
Reprint requests to Andrea Frustaci, Istituto di Cardiologia, Università Cattolica del Sacro Cuore, Largo Gemelli 8, 00168 Roma, Italy.
- Received November 4, 1996.
- Revision received March 4, 1997.
- Accepted March 7, 1997.
- Copyright © 1997 by American Heart Association
Aretz H, Billingham ME, Edwards WD, Factor SM, Fallon JT, Fenoglio JJ, Olsen EGJ, Schoen F. Myocarditis: a histopathologic definition and classification. Am J Cardiovasc Pathol. 1986;1:3-14.
Daly K, Richardson PJ, Olsen EGJ, Morgan-Capner P, McSorley C, Jackson G, Jewitt DE. Acute myocarditis: role of histological and immunological examination in the diagnosis and the assessment of immunosuppressive treatment. Br Heart J. 1984;51:30-35.
Morillo CA, Klein GJ, Jones DL, Guiraudon CM. Chronic rapid atrial pacing: structural, functional, and electrophysiological characteristics of a new model of sustained atrial fibrillation. Circulation. 1995;91:1588-1595.
Fitchett DH, Sugrue DD, MacArthur CG, Oakley CM. Right ventricular dilated cardiomyopathy. Br Heart J. 1984;51:25-29.
Sekiguchi M, Hiroe M, Kasanuki H, Ohnishi S, Hirosawa K. Experience of 100 atrial endomyocardial biopsy and the concept of atrial cardiomyopathy. Circulation. 1984;70(suppl II):II-118.
James TN. Diversity of histopathologic correlates of atrial fibrillation. In: Kulbertus HE, Olsson SB, Schlepper M, eds. Atrial Fibrillation. Molndal, Sweden: AB Hassle; 1982:13-30.
Baandrup U, Florio RA, Olsen EGJ. Do endomyocardial biopsies represent the morphology of the rest of the myocardium? A quantitative light microscopic study of single v multiple biopsies with the King’s bioptome. Eur Heart J. 1982;3:171-178.
Fromer M, Genton C, Schlaepfer J, Goy JJ, Kappenberger L. Is there an isolated arrhythmogenic right atrial myocarditis? Eur Heart J. 1990;11:566-571.
Balaji S, Wiles HB, Sens MA, Gillette PC. Immunosuppressive treatment for myocarditis and borderline myocarditis in children with ventricular ectopic rhythm. Br Heart J. 1994;72:354-359.