An Arrhythmogenic Substrate for Persistent Atrial Fibrillation
Background— Structural changes, like atrial fibrosis, may increase the likelihood of atrial fibrillation (AF) occurring in response to triggering events. The influence of isolated atrial amyloidosis (IAA) is largely unknown.
Methods and Results— Right atrial appendages (1 or 2 entire cross sections) were obtained from 245 patients undergoing open-heart surgery. Atrial amyloid was identified by Congo red staining and classified by immunohistochemistry. Amyloid was found in 40 (16.3%) of 245 patients, and all deposits were immunoreactive for atrial natriuretic peptide (ANP). Thirty-eight (15.5%) patients suffered from persistent AF. The presence of amyloid correlated with age and P-wave duration and was related to sex, valve diseases, and the presence of AF (P<0.01). The association between atrial amyloid, AF, and P-wave duration was independent of age and sex. According to multiple logistic regression analysis, amyloid was the only age- and sex-independent predictor for the presence of AF. Atrial fibrosis was not a predictor for AF, and the amount of amyloid correlated inversely with the degree of interstitial fibrosis (P=0.001; r=−0.55).
Conclusions— Our study provides evidence that IAA affects atrial conduction and increases the risk of AF. The occurrence of IAA depends on age leading to the formation of an amyloid nidus. The progression and consequences of IAA are then influenced by pathological conditions, such as valve diseases, that increase synthesis and secretion of ANP. The inverse correlation between IAA and atrial fibrosis suggests that these patients may not benefit from treatment with ACE inhibitors to reduce the amount of atrial fibrosis.
Received June 19, 2002; revision received August 7, 2002; accepted August 7, 2002.
Previous studies have shown that atrial fibrosis provides a structural substrate for AF.1 Areas of fibrotic tissue cause conduction inhomogeneities and are included in macro reentry circuits during AF.2 Increased amounts of atrial fibrosis were found in patients with AF, who also showed an activated angiotensin system.3,4⇓ In addition to angiotensin II–related changes, the amount of atrial fibrosis increases with patient age.4
Another structural change commonly observed with increasing age in the heart is amyloidosis, which, like atrial fibrosis, may disturb atrial conduction. Amyloidosis represents a diverse group of diseases characterized by the presence of extracellular proteinaceous deposits showing characteristic structural and tinctorial properties. Amyloidoses are classified based on the fibril protein deposited and the clinical presentation. They affect the heart as part of systemic amyloidosis as in AL amyloidosis, which is caused by the deposition of a portion of immunoglobulin light chains. Amyloid may also be deposited in the heart as a manifestation of aging. In senile cardiovascular amyloidosis, the fibril protein consists of transthyretin and amyloid is observed in cardiac vessels and the interstitium of the ventricles and atria. Most commonly, the heart is affected by a strictly localized or organ-limited variant called isolated atrial amyloidosis (IAA). The incidence of IAA increases with age, reaching >90% in the ninth decade.5–8⇓⇓⇓ The fibril protein deposited in IAA is atrial natriuretic peptide (ANP), a peptide hormone synthesized and secreted predominantly by atrial cardiomyocytes.9,10⇓
Although IAA is far more common than AL amyloidosis or senile cardiovascular amyloidosis, little is known about the putative role of IAA in the pathogenesis of cardiac arrhythmias in the elderly.11 We used right atrial appendages to study the relationship between the presence of atrial amyloidosis and the prevalence of persistent AF. We believe that this is the first large-scale study to investigate the potential contribution of atrial amyloidosis in relation to previously reported factors, such as atrial conducting abnormalities and fibrosis, on the occurrence of AF.
Right atrial appendages were obtained from 245 patients undergoing cardiac bypass surgery, mitral/aortic valve replacement, and other surgical procedures at the Department of Cardiothoracic Surgery, University Hospital Magdeburg (Table 1). Patients included in the study had to be older than 18 years of age and scheduled for elective open-heart surgery requiring cardiopulmonary bypass. Patients were excluded if they had paroxysmal atrial fibrillation, if they had previously participated in another investigational protocol, or if they had uncontrolled renal, hepatic, or heart failure or a known history of AL and transthyretin-derived (ATTR) amyloidosis. All patients gave written consent to participate in the study, and their baseline characteristics are shown in Table 1. The investigation conforms with the principles outlined in the Declaration of Helsinki.
All patients were hospitalized 1 day before surgery. After clinical examination, a 12-lead ECG was recorded for each patient, and routine blood samples were taken to determine white and red blood cell counts, hepatic and renal function, and C-reactive protein level. All surgical procedures were performed using extracorporal circulation. During the surgery, every patient received myocardial protection by cold cardioplegia (Bretschneider solution). After the procedure, all patients were monitored according to common clinical practice. The P-wave duration was assessed preoperatively using manual measurements from 2 independent observers in lead II from a standard 12-lead surface ECG (50 mm/sec paper speed).
The patient groups were formed as follows: group 1 included all 245 study patients and group 2 was obtained from group 1 by matching 40 patients with amyloid by age and sex with 40 control patients without amyloid. Group 3 was obtained from group 1 by matching 38 patients with persistent AF by age and sex with 38 control patients in sinus rhythm (SR).
Histochemistry and Immunohistochemistry
Tissue samples (1 or 2 entire cross sections, depending on the amount of tissue available) of the right atrial appendages were fixed in 10% buffered formalin and embedded in paraffin. Deparaffinized sections were stained with H&E and van Gieson’s elastic stain (EvG). The presence of amyloid was demonstrated by the appearance of green birefringence from alkaline alcoholic Congo red staining under polarized light. Immunostaining was performed as described elsewhere12 using the following antibodies: monoclonal antibody directed against AA amyloid (dilution 1:500), polyclonal antibodies directed against atrial natriuretic peptide (1:1000; BioGenesis, England), transthyretin (1:600), β2-microglobulin (1:2000), λ-light chain (1:7500), and κ-light chain (1:7500) (all from DAKO, Glostrup, Denmark). The specificity of immunostaining was controlled using specimens containing known classes of amyloid. Omission of primary antibodies served as negative controls.
Quantitation of Atrial Fibrosis and Amyloid
The volume percentage (V%) of atrial fibrosis, including interstitial and perivascular fibrosis, was quantitatively analyzed with point counting on EvG-stained paraffin sections, as described elsewhere.13 Quantitative analysis of amyloid was performed using Congo red fluorescence. Congo red–stained specimens were scanned entirely using a fluorescence microscope (Olympus AH-3) equipped with 100-W high-pressure mercury burner and fluorescein isothiocyanate filter.14 The number of high-power fields (at ×40 objective) per specimen with amyloid was divided by the number of high-power fields covered by the entire specimen and given as amyloid index (AI). Point counting was not feasible, because the grid was not discernible in fluorescence microscopy. The pathologist (C.R.) who read the biopsy specimens and quantitated the amount of fibrosis and amyloid was unaware of the personal and clinical data of the patients.
All values are expressed as mean±SD. Continuous variables were compared by means of the unpaired Student’s t test. The χ2 test, multivariance analysis, and principal-component multivariance method were used, where appropriate, to assess the association between the occurrence of AF, amyloid, and clinical parameters. The Pearson correlation coefficient was used to determine the relationship between metric parameters. A value of P<0.05 was considered to be statistically significant.
Table 1 summarizes the clinical characteristics of the patients. The mean age of the patients was 63.1±10.4 years (range, 24 to 79 years), including 174 men and 71 women.
Amyloid was found as interstitial, perivascular, or endocardial deposits in 40 (16%) patients (Figure 1; Table 1). The mean age of the patients with amyloid was 69.6±7.6 years (range, 39 to 79 years). The amount of amyloid ranged from 0.06 to 1.00 AI (mean, 0.35±0.26 AI). The amyloid deposits were immunoreactive for ANP in all 40 (100%) cases. Four (10%) cases showed additional transthyretin-positive amyloid deposits. No other antibody stained amyloid.
Atrial fibrosis was characterized by progressive thickening of the interstitial matrix initially involving groups of myocytes, subsequent separation of individual myocytes, and, finally, formation of large patches of confluent fibrotic areas. The volume percentage of atrial fibrosis ranged from 5.0% to 23.5% (13.5±3.6%; Table 1).
Thirty-eight (16%) of the 245 patients studied suffered from persistent AF (Table 1); their mean age was 68.4±7.2 years (range, 49 to 79 years).
Entire Patient Population
Table 1 summarizes the univariate analyses calculated from all study patients (group 1). Amyloid was found more commonly in the atria of patients with AF than in those in SR (P<0.01). However, both amyloid and AF were found more commonly in women (P<0.01), with increasing age (P<0.01; Table 1), and in patients undergoing mitral valve replacement (P<0.01). Interestingly, the amount of amyloid correlated inversely with the amount of interstitial fibrosis (P=0.001; r=−0.55; Figure 2), whereas right atrial fibrosis was not related to the presence of AF or type of heart surgery. According to multiple logistic regression analysis, patient age, P-wave duration, and type of heart surgery were independent predictors for the presence of amyloid, whereas patient age, coronary artery disease, and type of heart surgery were independent predictors of AF.
Matched Patient Populations
Group 2 (amyloid versus nonamyloid; Table 2) and group 3 (AF versus SR; Table 3) included patient populations matched by age and sex. Tables 2 and 3⇓ summarize the results of univariate analyses. The relationship between atrial amyloid and AF proved to be independent of age and sex. Patients with amyloid were significantly more likely to suffer from AF (P<0.01; Table 2) than those without amyloid, and patients with AF had larger amounts of amyloid in their appendages than patients in SR (P=0.04; Table 3). The presence of AF also depended on the type of heart surgery and was more common in patients undergoing mitral valve replacement (Tables 2 and 3⇓). Again, atrial fibrosis was not related to the occurrence of AF, whereas the amount of fibrosis correlated inversely with the amount of amyloid in group 2 (P=0.001; r=−0.55).
According to multiple logistic regression analysis, both calcium channel blockers and P-wave duration predicted the presence of amyloid in group 2 (amyloid versus nonamyloid), whereas in group 3 (AF versus SR), P-wave duration was the only predictor of the presence of amyloid. In group 2 (amyloid versus nonamyloid), the presence of amyloid and treatment with digitalis was associated with AF, but in group 3 (AF versus SR), only the amount of amyloid showed this association. Thus, after adjustment for age and sex, amyloid, not interstitial fibrosis, remains the only predictor of AF in patients undergoing mitral valve replacement.
Isolated Atrial Amyloidosis
Isolated atrial amyloidosis (IAA) belongs to the family of senile amyloidoses, and as observed in our series, the incidence of IAA increases with age.5–8⇓⇓⇓ The atrial amyloid deposits of our patients were immunoreactive for ANP and are thus interpreted as IAA. Four patients also showed vascular and interstitial amyloid deposits immunoreactive for transthyretin. None of these patients had clinical evidence of polyneuropathy or gave a family history of amyloidosis and were thus interpreted as senile cardiovascular amyloidosis. Simultaneous deposition of IAA and senile cardiovascular amyloidosis in the atria has been described previously.7
Pathogenesis of Isolated Atrial Amyloidosis
The pathogenesis and clinical consequences of IAA are poorly defined. In general, the formation of amyloid fibrils takes place by a nucleated growth mechanism. In the absence of an aggregate, a nucleus must be generated in situ in a process that is dependent on protein concentration and requires significant time delays. Once a nucleus is formed, rapid postnucleation fibril growth occurs following a first order kinetic model with no lag time. Endogenous accessory factors affect the rate of amyloid formation by promoting or inhibiting the aggregation process. There is evidence to suggest that this mechanism of aggregation and fibril formation also accounts for IAA.15 As IAA is formed locally at the site were ANP is synthesized, a high local concentration of the precursor protein contributes to the formation of amyloid,16 which means that cardiac and extracardiac stimuli of synthesis and secretion of ANP may support amyloidogenesis.
Atrial Natriuretic Peptide and Isolated Atrial Amyloidosis
ANP plays an integral role in the regulation of hydromineral homeostasis under normal and pathological conditions. ANP secretion is stimulated by volume expansion, supine posture, tachycardia, exercise, hypoxia, myocardial ischemia, and other factors.17 Plasma ANP levels are raised in heart failure in proportion to the severity of cardiac dysfunction, irrespective of the etiology of ventricular failure. There is a positive correlation between ANP level, cardiac filling pressure, and renin level.18 Serum ANP levels are significantly higher in elderly patients; however, this may be related to the increased incidence of cardiac diseases, rather than resulting from aging alone.19
In our series, the presence of amyloid correlated with age, sex, P-wave duration, persistent AF, and type of heart surgery. Amyloid did not correlate with NYHA functional class or ejection fraction, indicating that not all cardiac conditions, which may increase ANP serum levels, promote atrial amyloidogenesis. The increased prevalence of amyloid in women in our series was related to age differences: the women were significantly older than men. However, Steiner6 and Hodkinson and Pomerance11 found atrial amyloid deposits more frequently in women than in men, the difference being most significant in the age group 31 to 50 years.6
The most interesting observation was the high prevalence of amyloid among patients undergoing mitral valve replacement (Table 1). Looi20 made a similar observation and found IAA significantly more common in patients with chronic rheumatic heart disease. Mitral valve disease may cause a significant dilation and hypertrophy of the left atrium, which in turn stimulates synthesis and secretion of ANP, finally contributing to the deposition of IAA. Mitral valve disease is commonly associated with AF.
IAA and AF: Epiphenomenon or Etiologic Link?
These observations raise the question of whether the pathogenesis of IAA is associated with AF. Multiple regression logistics showed that age, coronary artery disease, and type of heart surgery were strong predictors of AF but not of amyloid in all study patients. The presence of amyloid correlated with patient age, P-wave duration, and type of heart surgery, indicating that AF is not a major risk factor for the development of amyloid. Indeed, IAA is far more common than AF: nearly 10% of the octogenarian population suffer from AF, 1 whereas the prevalence of IAA reaches 90%.5–8⇓⇓⇓ However, after matched patient populations were selected in our study, it became evident that AF is associated with significantly higher amounts of amyloid compared with SR, suggesting that AF modulates the amount of amyloid deposited rather than being an initiator of amyloid formation.
Several studies have found significantly raised serum levels of ANP in AF in general19,21–23⇓⇓⇓ and in mitral valve diseases in particular.24,25⇓ The mRNA content of ANP is increased in patients with valvular disease,25 additionally substantiating the hypothesis that the positive correlation between IAA, type of heart surgery, and AF is based on an increased synthesis and secretion of ANP accelerating the deposition of amyloid. In animal models, ANP levels increase after AF has been induced.26,27⇓ Thus we hypothesize that the occurrence of IAA is influenced by age and other yet unknown variables leading to the formation of an amyloid nidus. The progression of IAA is then influenced by diseases affecting the heart, such as mitral valve disease and persistent AF. In 1977, Hodkinson and Pomerance11 described a correlation between senile cardiac amyloidosis and AF. However, their study provided no information about the type of fibril protein deposited and did not consider the influence of age and sex.11
Consequences of Amyloid Deposition
Once amyloid is deposited, it is a permanent structural alteration of the atria. ANP fibrils induce apoptosis,28 and amyloid deposited in the heart disturbs myocyte contractility and conduction. The P-wave duration was significantly longer in patients with amyloid, and it was an independent predictor in all 3 patient groups, indicating that amyloid influences atrial conduction. We believe that our study provides evidence that the pathology of IAA is associated with AF and that IAA affects atrial conduction. In fact, according to multiple logistic regression analysis, amyloid was the only parameter associated with AF after adjustment for age and sex. Thus, the presence of amyloid enhances the susceptibility for AF. AF then increases the amount of amyloid deposited, leading to a vicious circle.
Recent studies have shown the importance of atrial fibrosis for the occurrence of AF. One molecular mechanism for the development of fibrosis is the atrial angiotensin system, and therapy with ACE inhibitors offers a novel therapeutic approach to AF.1,29⇓ However, we observed a significant inverse correlation between the amount of amyloid deposited and the degree of interstitial fibrosis; this is supported by the study of Yoshihara et al, 30 in which plasma ANP levels correlated inversely with left atrial collagen volume. Indeed, in our series, fibrosis had no impact on AF, and we may have found a patient population that does not benefit from ACE inhibitors: patients with mitral and aortic valve disease or advanced IAA.
Interestingly, an inverse correlation was found between amyloid and calcium channel blockers in our series. The reason for this relationship remains unclear, and additional studies including larger groups of patients are warranted to investigate the potential impact of calcium channel blockers on amyloid deposition.
Some potential limitations may have influenced our results. All patients included in our study underwent cardiac surgery. Thus, no comment can be made about the possible impact of amyloid in other patient populations. Because left atrial appendages were not analyzed, no comment can be made about the influence of interatrial distribution of IAA on AF. It is necessary to sample larger amounts of human atrial tissue to answer these questions, but this procedure is not feasible. To demonstrate a statistically significant association between amyloid, age, P-wave duration, and the risk of AF, it presumably suffices to characterize right atrial amyloidosis. Electron microscopy was not performed, and atrial size was not assessed systematically. Thus, no comment can be made about the ultrastructural appearance and the relation of atrial amyloidosis to atrial size. In addition, future studies will have to investigate the impact of IAA on interventional strategies like catheter ablation of focal AF or surgical maze procedure.1,30⇓
- ↵Li D, Fareh S, Leung TK, et al. Promotion of atrial fibrillation by heart failure in dogs: atrial remodeling of a different sort. Circulation. 1999; 100: 87–95.
- ↵Davies MJ, Pomerance A. Pathology of atrial fibrillation in man. Br Heart J. 1972; 34: 520–525.
- ↵Kaye GC, Butler MG, D’Ardenne AJ, et al. Isolated atrial amyloid contains atrial natriuretic peptide: a report of six cases. Br Heart J. 1986; 56: 317–320.
- ↵Hodkinson HM, Pomerance A. The clinical significance of senile cardiac amyloidosis: a prospective clinico-pathological study. Q J Med. 1977; 183: 381–387.
- ↵Goette A, Jünemann G, Peters B, et al. Determinants and consequences of atrial fibrosis in patients undergoing open heart surgery. Cardiovasc Res. 2002; 54: 390–396.
- ↵Wallen T, Landahl S, Hedner T, et al. Atrial peptides, ANP(1–98) and ANP(99–126) in health and disease in an elderly population. Eur Heart J. 1993; 14: 1508–1513.
- ↵Berglund H, Boukter S, Theodorsson E, et al. Raised plasma concentrations of atrial natriuretic peptide are independent of left atrial dimensions in patients with chronic atrial fibrillation. Br Heart J. 1990; 64: 9–13.
- ↵Wijffels MC, Kirchhof CJ, Dorland R, et al. Electrical remodeling due to atrial fibrillation in chronically instrumented conscious goats: roles of neurohumoral changes, ischemia, atrial stretch, and high rate of electrical activation. Circulation. 1997; 96: 3710–3720.
- ↵Schubert D, Behl C, Lesley R, et al. Amyloid peptides are toxic via a common oxidative mechanism. Proc Natl Acad Sci U S A. 1995; 92: 1989–1993.
- ↵Li D, Shinagawa K, Pang L, et al. Effects of angiotensin-converting enzyme inhibition on the development of the atrial fibrillation substrate in dogs with ventricular tachypacing-induced congestive heart failure. Circulation. 2001; 104: 2608–2614.
- ↵Yoshihara F, Nishikimi T, Sasako Y, et al. Plasma atrial natriuretic peptide concentration inversely correlates with left atrial collagen volume fraction in patients with atrial fibrillation: plasma ANP as a possible biochemical marker to predict the outcome of the maze procedure. J Am Coll Cardiol. 2002; 39: 288–294.