Feasibility and Safety of Atrial Biopsy　― Evidence From 1,000 Cases ―

Abstract

Background

Atrial biopsy is technically challenging owing to the atria’s thin walls and relatively thick endocardium. This study assessed the feasibility and safety of echocardiography-guided atrial biopsy in a consecutive cohort of 1,000 patients who underwent catheter ablation for atrial tachyarrhythmias or percutaneous left atrial (LA) appendage occlusion.

Methods and Results

Atrial biopsy was performed at the limbus of the fossa ovalis through the femoral vein using a 5.5-Fr (n=233) or a 7.0-Fr (n=767) bioptome under intracardiac (n=963) or transesophageal (n=37) echocardiography guidance, alongside fluoroscopy. For histological analysis, 5 tissue samples were collected from the same site. Biopsy was successfully completed in 996 (99.6%) patients. Patients were divided based on histological depth into Group A (biopsy beyond the endocardium; n=885) and Group B (endocardial-only biopsy; n=111). Multivariable logistic regression identified larger LA volume, use of a 5.5-Fr bioptome, and amyloid deposition as independent predictors of Group B (P=0.009, P<0.001, and P=0.001, respectively). Moreover, biopsy-related complications were unrecorded.

Conclusions

Echocardiography-guided atrial biopsy is a feasible and safe technique. However, atrial enlargement, smaller bioptome size, and amyloid deposition are associated with unsuccessful endocardial penetration and collection of myocardial tissue.

Atrial cardiomyopathy influences the development of atrial tachyarrhythmias, heart failure, and stroke.^1–3 However, histological evaluation of atrial tissue has been limited by the difficulty of obtaining atrial samples. Consequently, the assessment of atrial cardiomyopathy has primarily relied on imaging and electrophysiological findings.^4–9 Although fibrosis has traditionally been considered the principal structural change, recent histological studies state that additional histopathological alterations, such as intercellular space expansion, severity of myofibrillar loss, reduction in myocardial nuclear density, and amyloid deposition, contribute to the progression of atrial cardiomyopathy, as reflected by atrial voltage reduction.^6,10–11

To the best of our knowledge, prior to our 2022 report on intracardiac echocardiography (ICE)-guided atrial biopsy in patients with atrial fibrillation,⁶ only 2 studies have reported relatively small numbers of atrial biopsy cases. One study, published in 1997, conducted biopsy under fluoroscopic guidance (n=12)¹² and the other, published in 2020, utilized transesophageal echocardiographic guidance (n=4).¹³ Our 2023 study revealed histological factors associated with atrial voltage based on 230 ICE-guided atrial biopsy cases.¹⁰ More recently, based on ICE-guided atrial biopsies, we reported a 7% prevalence of atrial amyloidosis in 578 patients.¹¹ In these reports, the biopsy technique was described in detail. Although these studies have primarily focused on specimens containing sufficient myocardial tissue, approximately 20% of the cases in our reports lacked adequate myocardial tissue for histological assessment, partly because of failure to penetrate the atrial endocardium. The endocardium of the right atrial septum – the target site for atrial biopsy – is relatively thick (mean 0.85 mm),¹⁰ which may hinder penetration and prevent the acquisition of sufficient myocardial tissue. In such cases, only the endocardial tissue is obtained, precluding histological assessment of the myocardial layer.

We evaluated the feasibility and safety of performing this technique in a consecutive cohort of 1,000 patients who underwent catheter ablation for atrial tachyarrhythmias or percutaneous left atrial appendage occlusion (LAAO) for the future clinical implementation of this atrial biopsy technique, identifying risk factors for unsuccessful endocardial penetration and collection of adequate myocardial tissue.

Methods

Patient Population

This study is a part of several ongoing prospective observational studies, such as the Histological Evaluation of Atrial Fibrillation Substrate Based on Atrial Septum Biopsy (HEAL-AF Study, HEAL-AF Study 2, and HEAL-AF Study 4), Follow-up Study of Patients Undergoing Catheter Ablation for Atrial Fibrillation: Evaluation of Long-term Outcomes and Predictive Factors Based on Genetic Predisposition–Subgroup Analysis in Patients with Atrial Biopsy and Proteomics (FUTURE-AF-S), Left Atrial Appendage Occlusion and Histological Assessment of Atrial Substrate in Patients with Atrial Fibrillation (LEARNMORE), and Histological Evaluation of Atrial Cardiomyopathy Based on Atrial Septum Biopsy in Patient Without Atrial Fibrillation (HISTORY). All patients underwent endomyocardial atrial biopsy between June 2020 and April 2025. Overall, 1,000 consecutive Japanese patients were enrolled, including 918, 42, and 36 who underwent AF ablation, supraventricular tachycardia ablation, and LAAO, respectively. Patient characteristics are summarized in Table 1. No patients were previously diagnosed with cardiac amyloidosis.

Table 1.

Patient Characteristics in the Entire Cohorts

Variable	Total (n=996*)	AF ablation (n=918)	Non-AF ablation (n=42)	LAAO (n=36)
Age (years)	67±12	67±11	54±17	75±6
Sex, female	307 (31)	277 (30)	20 (48)	10 (28)
BMI (kg/m2)	24±4	25±4	24±5	24±3
AF type
Paroxysmal AF	395 (41)	388 (42)	0 (0)	7 (19)
Non-paroxysmal AF	558 (59)	529 (58)	0 (0)	29 (81)
Hypertension	580 (58)	536 (58)	14 (33)	30 (83)
Diabetes	172 (17)	162 (18)	2 (5)	8 (22)
History of congestive heart failure	246 (25)	239 (26)	1 (2)	6 (17)
eGFR (mL/min/1.73 m2)	62±17	62±17	79±14	57±13
Previous PVI	128 (13)	119 (13)	0 (0)	9 (25)
LA diameter (mm)	41±7	41±6	33±5	44±6
LA volume/BSA (mL/m2)	44±16	46±24	28±9	56±20
Type of anesthesia during ablation
General anesthesia	281 (28)	245 (27)	0 (0)	36 (100)
Deep sedation	715 (72)	673 (73)	42 (100)	0 (0)
Type of sheath
Non-steerable sheath	166 (17)	136 (15)	0 (0)	30 (83)
Steerable sheath	828 (83)	780 (85)	42 (100)	6 (17)
Bioptome size
5.5 Fr (2.46 mm3)	233 (23)	233 (25)	0 (0)	0 (0)
7.0 Fr (5.20 mm3)	763 (77)	688 (75)	42 (100)	37 (100)
Histological parameter
%Fibrosis (%)	6 [4–9]	6 [4–9]	5 [4–9]	13 [8–19]
%Intercellular space (%)	27 [21–32]	27 [22–33]	19 [16–26]	22 [17–28]
%Myofibrillar loss (%)	15 [11–20]	15 [11–21]	12 [9–16]	15 [11–20]
Nuclear density (/mm2)	525 [403–680]	532 [415–684]	615 [483–745]	279 [194–351]
Amyloid deposition	80 (8)	74 (8)	1 (2)	5 (14)
Type of amyloid (n)
ATTR	45	41	1	3
AL	6	6	0	0
ANP	15	15	0	0
Undetermined	14	12	0	2

*Four patients in whom biopsy was unsuccessful were excluded. Unless indicated otherwise, data are presented as n (%), median [IQR] or mean±SD. AF, atrial fibrillation; AL, amyloid light chain; ANP, atrial natriuretic peptide type; ATTR, amyloid transthyretin; BMI, body mass index; BSA, body surface area; eGFR, estimated glomerular filtration rate; LA, left atrial; LAAO, left atrial appendage occlusion; PVI, pulmonary vein isolation.

The study protocols were approved by the Ethics Committee of Saga University Hospital (approval reference numbers: 20200101 for HEAL-AF, 20200901 for HEAL-AF2, 202200401 for FUTURE-AF-S, 20240804 for HEAL-AF4, 20230203 for HISTORY, and 20230501 for LEARNMORE). Written informed consent was obtained from all participants. This study was conducted in accordance with the principles of the Declaration of Helsinki.

Endomyocardial Atrial Biopsy

Endomyocardial atrial biopsy and catheter procedures were performed in the post-absorptive state under general anesthesia (n=283; 28%) or deep sedation using midazolam, dexmedetomidine hydrochloride, and fentanyl (n=717; 72%). In all cases, the femoral vein was used for venous accesses. Atrial biopsy was performed before transseptal puncture in all patients. In 964 patients, ICE (ViewFlex^TM, Abbott, St Paul, MN, USA) was used in the right atrium (RA) to visualize the long-axis view of the atrial septum, including the fossa ovalis (FO) and the limbus of the FO (Figure 1). In 36 patients, transesophageal echocardiography (TEE, EPIQ CVx Ultrasound System with an X8-2t Probe; Philips, Amsterdam, The Netherlands) was used (Figure 1). A bioptome (5.5 Fr, 104 cm, 2.46 mm³ tip volume, or 7.0 Fr, 104 cm, 5.20 mm³ tip volume; Cordis, Miami Lakes, FL, USA) was advanced to the right atrial septum through a steerable sheath (Agilis^TM, Abbott; n=834; 83%) or a conventional non-steerable sheath (Swartz SL0^TM; Abbott; n=166; 16%). Biopsy was performed at the limbus of the FO. Additionally, fluoroscopy was used to confirm the approximate location of the bioptome and its opening in the left anterior oblique (LAO) view (Figure 1). Previously, we demonstrated that the atrial endocardium thickness in the region corresponding to the biopsy site averaged 0.85±0.15 mm in autopsy cases.¹⁰ Owing to this endocardial thickness, the initial biopsy typically obtained only grossly white endocardial tissue (Figure 1). To increase the likelihood of obtaining myocardial tissue, which grossly appears red, subsequent biopsies were repeatedly directed at the same site (Figure 1). In some cases, ICE or TEE revealed the formation of a visible recess at the repeated biopsy sites (Figure 2).¹¹ Biopsy sites were continuously monitored using ICE or TEE, and if the recess depth exceeded approximately 3 mm, no further biopsies at that site were performed. Additionally, if biopsy samples floated in 4% paraformaldehyde solution, suggesting that the biopsy had reached the fat layer deep within the atrial septum, the biopsy at that site was also terminated, and subsequent biopsies were performed approximately 5–10 mm away from the original site.¹¹ Occasionally, remnants of the left venous valve, appearing as single or multiple strands at the limbus of the FO, were obtained during biopsy, particularly when sampling was performed at the lower margin of the limbus.¹¹ During the subsequent ablation, biopsy sites were monitored using ICE or TEE. Five tissue samples measuring 1–3 mm were obtained from each patient and immediately fixed in 4% paraformaldehyde solution.

Figure 1.

Intracardiac echocardiography (ICE)- or transesophageal echocardiography (TEE)-guided atrial biopsy. Fluoroscopy was also used to roughly guide the bioptome during catheter ablation (A). Representative ICE image of the limbus of the fossa ovalis (FO; B). Representative biopsy samples (C). The white arrows indicates a sample consisting only of endocardial tissue, whereas the other samples appear to contain myocardial tissue. Fluoroscopic image of the bioptome and TEE probe during left atrial appendage occlusion viewed from the left anterior oblique (LAO) projection (D). Biplane TEE images during atrial biopsy showing the bioptome tenting the limbus of the FO (E,F). Ao, aorta; CS, coronary sinus; LA, left atrium; RA, right atrium; SVC, superior vena cava.

Figure 2.

An example of an intracardiac echocardiography (ICE) image of the atrial septum during atrial biopsy. The ICE image obtained after the initial biopsy is shown (A), where only a slight (shallow) recess was observed at the biopsy site (arrowhead). After repeated biopsies at the same site, a more pronounced recess (arrowhead) was observed (B). FO, fossa ovalis; LA, left atrium.

Complications of the Atrial Biopsy

Biopsy-associated complications include cardiac perforation, acute swelling at the biopsy site likely due to hematoma, and atrioventricular block. Acute swelling of the biopsy site was defined as swelling occurring before the application of radiofrequency or cryoenergy to the right pulmonary veins, as these thermal energy applications can cause acute swelling of the atrial septum (Supplementary Figure). Chronic-phase imaging assessments were performed using transthoracic echocardiography at 1 year after the procedure in ablation patients and TEE at 4 months in LAAO cases.

Histological Assessment

Biopsy samples were fixed in 4% paraformaldehyde for at least 48 h and subsequently embedded in paraffin. Tissue sections were cut at a thickness of 5 µm, and the deparaffinized sections were stained with hematoxylin and eosin, Masson’s trichrome, and Congo red. All staining procedures were performed according to a standardized protocol. Each stained slide was scanned using a digital slide scanner (NanoZoomer S60, Hamamatsu, Japan). Sufficient myocardial tissue samples were defined as those with a total analysis area ≥100,000 μm².¹⁰

Details of the histological quantitative analysis have been described elswhere.¹¹ Briefly, the quantitative analysis was performed using the HALO® and HALO AI image analysis platform (version 4.0.5107; Indica Labs, Albuquerque, NM, USA). Initially, the myocardium was manually selected in Masson’s trichrome-stained samples, excluding areas with crush artifacts caused by the bioptome and regions containing large blood vessels. The analysis area was subsequently defined using a first classifier trained with HALO AI’s DenseNet AI V2 algorithm. This classifier excluded the slide glass, atrial endocardium, small blood vessels, and perivascular interstitium, selecting only the myocardium and interstitial spaces between cardiomyocytes (endomysial space). Then, the myocardium and interstitium within the analysis area were segmented using a second classifier. The segmentation results were reviewed visually, and any remaining artifacts or vascular components in the analysis area were excluded manually before reapplying the classifiers.

The extent of fibrosis (%Fibrosis), intercellular space (%Intercellular space), and myofibrillar loss (%Myofibrillar loss) were measured (Figure 3). Fibrosis was assessed by highlighting the blue-stained areas from Masson’s trichrome sections in red and calculating the ratio of the red-stained area to the total analysis area. The extent of %Intercellular space was calculated by subtracting the fibrotic area from the interstitial space and dividing by the total analysis area. Myofibrillar loss was defined by referencing the darkest red-purple-stained areas of myocytes. Areas exhibiting staining intensity <50% of this reference were classified as myofibrillar loss, and the percentage of this area relative to the total myocardial area was calculated. Myocardial nuclear density was manually assessed through hematoxylin and eosin staining by counting the number of myocardial nuclei per analysis area as a surrogate for myocyte number. Myocarditis was diagnosed according to the Dallas criteria, defined as the presence of a myocardial inflammatory infiltrate with necrosis and/or degeneration of adjacent cardiomyocytes of non-ischemic nature.¹⁴ Amyloid typing was performed in cases with positive Congo red staining and apple-green birefringence under polarized light.^10,11 Details of the amyloid typing have been described elsewhere.¹¹

Figure 3.

Histological quantification using the HALO AI image analysis platform based on hematoxylin and eosin staining (A,B), Masson’s trichrome staining (C), and Congo red staining (D). The area outlined by the white square in (A) is shown at higher magnification in (B). (C) and (D) represent images of the corresponding area in (B) on the serial sections. The HALO AI algorithm selected only the myocardium and the interstitial spaces between the cardiomyocytes. The extent of endomysial fibrosis (%Fibrosis) was then measured (E). Blue-stained areas indicating fibrosis on Masson’s trichrome staining (C) are highlighted in red in the fibrosis map (E). The extent of the intercellular space (%Intercellular space) was calculated by subtracting the fibrotic area from the interstitial space (highlighted in green; F) and dividing by the total analysis area. Myofibrillar loss was characterized by regions with contrast intensity <50% of the reference value and expressed as a percentage of the total myocardial area (%Myofibrillar loss; G). Myocardial nuclear density was manually assessed on hematoxylin and eosin-stained sections by counting the number of myocardial nuclei per analysis area (yellow circles; H). Histological quantification methods are described in the main text. Refer to the scale bars in each panel. In this case, %Fibrosis, %Interstitial space, %Myofibrillar loss, and myocardial nuclear density were 19.1%, 33.9%, 11.9%, and 305 nuclei/mm², respectively. %Intercellular space is calculated by subtracting %Fibrosis from %Interstitial space, resulting in 14.8%.

Based on the histological depth of the biopsy samples, patients were divided into Group A, wherein the biopsy penetrated beyond the endocardium (biopsy beyond the endocardium), and Group B, wherein the biopsy did not penetrate the endocardium and only endocardial tissue was obtained (endocardial-only biopsy). Specifically, Group A involved cases in which sufficient myocardial tissue (defined as a total analysis area ≥100,000 μm²) was obtained or the fat layer was reached (Figure 4). Group B included cases in which sufficient myocardial tissue was not obtained. Group A was further subdivided into Group A1 (sufficient myocardial tissue group) and Group A2 (fat layer without myocardial tissue group; Figure 4).

Figure 4.

Classification of biopsy samples based on histological depth. Representative cases from Group A, wherein the biopsy penetrated beyond the endocardium, are shown (A,B). In (B), a large sample of fat tissue was obtained along with samples containing sufficient myocardial tissue (B, top panel). (A) and (B) were classified as Group A1 (sufficient myocardial tissue group). In contrast, in (C), only endocardial and fat tissue were obtained without sufficient myocardial tissue – defined as a total analysis area ≥100,000 μm² – resulting in classification as Group A2 (fat layer without myocardial tissue group). (D) A representative case from Group B, wherein the biopsy did not penetrate the endocardium and only endocardial tissue was obtained (endocardial-only biopsy). Notably, no myocardial tissue was obtained in this case. The bottom panels are the areas outlined by the white squares in the top panels, shown at higher magnification.

Statistical Analysis

Normally distributed data are expressed as the mean±standard deviation, whereas non-normally distributed data are presented as the median and interquartile range. Continuous variables were analyzed using the unpaired t-test for normally distributed data and the Wilcoxon rank-sum test for non-normally distributed data. Categorical variables were analyzed using the Chi-square test or Fisher’s exact test, as appropriate. A P value of <0.05 was considered significant. For comparisons among the 3 groups, the Bonferroni correction was applied, and a P value of <0.017 was considered significant. Univariable and multivariable logistic regression analyses were performed to identify clinical factors associated with unsuccessful endocardial penetration and collection of only endocardial tissue. The multivariable analyses included variables with a P value <0.05 in the univariable analyses. Statistical analyses were conducted using JMP® Pro (version 17.2; SAS Institute Inc., Cary, NC, USA).

Results

Atrial Biopsy Procedure

Patient characteristics are shown in Table 1. Biopsy procedures were performed by 5 physicians. Atrial tissue samples were successfully obtained within 10 min in 996 (99%) patients, excluding 4 in whom atrial biopsy could not be performed due to anatomical reasons, such as (1) scoliosis, (2) severe compression of the atrial septum by the elongated aortic root, (3) a potential risk of perforation due to thinning of the limbus of the FO, and (4) a markedly enlarged RA, resulting in the inability to advance the bioptome to the limbus of the FO using the SL0 sheath. A 5.5-Fr bioptome was used for the first 233 (23%) patients. After confirming the safety of the atrial biopsy technique, a 7.0-Fr bioptome was used in the subsequent 767 (77%) patients to obtain larger tissue samples. The results of the histological quantification are presented in Table 1. Amyloid deposition was identified in 80 (8%) patients. Amyloid phenotyping of the atrial samples identified transthyretin type (ATTR) in 45 patients, light chain type (AL) in 6 patients, all of whom had the lambda type, and atrial natriuretic peptide type (ANP) in 15 patients. The amyloid type could not be determined in 14 patients due to a minimal amount of amyloid deposits. An example of amyloid deposition and distribution of the amyloid type is shown in Figure 5. In all ATTR cases, amyloid deposition was consistently observed in at least the endocardium. No cases met the Dallas criteria for myocarditis.

Figure 5.

Examples of amyloid deposition in atrial samples. Congo red staining and polarized light microscopy showed amyloid deposition with characteristic apple-green birefringence observed in both the endocardium and myocardium. The dotted line indicates the boundary between the endocardium and myocardium. (A,B). (C,D) A case from Group B (the endocardial-only group) is shown; only the endocardium was obtained and amyloid deposition was observed within the endocardial layer. Amyloid deposition in atrial samples was identified in 80 (8%) of 996 patients (E). The distribution of each amyloid type is shown (F). AL, amyloid light chain; ANP, atrial natriuretic peptide; ATTR, amyloid transthyretin.

Histological Depth of the Biopsy Samples

Patients were divided into 2 groups based on the histological depth of the biopsy samples: 885 (89%) patients were categorized into Group A (biopsy beyond the endocardium) and 111 (11%) into Group B (endocardial-only biopsy). Univariable and multivariable logistic regression analyses identified larger atrial volume, use of a 5.5-Fr bioptome, and amyloid deposition as independent predictors of Group B (P: 0.009, <0.001, and 0.001, respectively; Table 2). In the subgroup of patients who underwent biopsy with a 7.0-Fr bioptome, only the presence of amyloid deposition was significantly more frequent in the endocardial-only biopsy group (P=0.02; Table 3).

Table 2.

Univariable and Multivariable Logistic Regression Analyses of Clinical Factors Associated With Endocardial-Only Biopsy

Variable	Univariate		Multivariate model
	OR (95% CI)	P value	OR (95% CI)	P value
Age	1.02 (0.99–1.03)	0.06
Female	1.19 (0.77–1.79)	0.41
BMI (kg/m2)	0.96 (0.91–1.00)	0.08
AF type
Paroxysmal AF	Ref.
Non-paroxysmal AF	1.41 (0.94–2.16)	0.09
Hypertension	0.76 (0.51–1.13)	0.18
Diabetes	0.79 (0.44–1.34)	0.39
History of congestive heart failure	1.14 (0.72–1.77)	0.55
eGFR (mL/min/1.73 m2)	0.99 (0.98–1.01)	0.53
LA diameter (mm)	1.04 (1.01–1.07)	0.01
LA volume/BSA (per 10 mL increase; mL/m2)	1.08 (1.01–1.16)	0.015	1.09 (1.03–1.17)	0.009
Type of sheath
Non-steerable sheath	Ref.
Steerable sheath	1.74 (0.97–3.41)	0.17
Bioptome size
5.5 Fr (2.46 mm3)	8.30 (5.46–12.8)	<0.001	9.52 (6.13–15.1)	<0.001
7.0 Fr (5.20 mm3)	Ref.
Amyloid deposition	2.00 (1.06–3.57)	0.03	3.34 (1.66–6.43)	0.001

CI, confidence interval, OR, odds ratio. Other abbreviations as in Table 1.

Table 3.

Univariable Logistic Regression Analyses of Clinical Factors Associated With Endocardial-Only Biopsy Among Patients Undergoing Atrial Biopsy Using a 7.0-Fr Bioptome

Variable	Univariate
	OR (95% CI)	P value
Age	1.02 (0.99–1.05)	0.24
Female	1.15 (0.56–2.25)	0.69
BMI (kg/m2)	0.96 (0.88–1.04)	0.29
AF type
Paroxysmal AF	Ref.
Non-paroxysmal AF	1.76 (0.88–3.74)	0.11
Hypertension	0.65 (0.34–1.23)	0.18
Diabetes	0.70 (0.44–1.67)	0.44
History of congestive heart failure	1.64 (0.81–3.18)	0.16
eGFR (mL/min/1.73 m2)	0.99 (0.97–1.01)	0.23
LA diameter, mm	1.03 (0.98–1.08)	0.24
LA volume/BSA (per 10 mL increase; mL/m2)	1.07 (0.98–1.14)	0.12
Type of sheath
Non-steerable sheath	Ref.
Steerable sheath	1.32 (0.58–3.56)	0.32
Amyloid deposition	2.80 (1.16–6.10)	0.02

Abbreviations as in Tables 1,2.

Among the 885 patients in Group A, 867 (97%) had sufficient myocardial tissue for histological analysis (Group A1: sufficient myocardial tissue group), whereas 18 (3%) had only endocardium and fat layer tissue (Group A2: fat layer without myocardial tissue group). Moreover, in the overall cohort (Supplementary Table 1) and subgroup of patients who underwent biopsy with a 7.0-Fr bioptome (Supplementary Table 2), no significant differences in clinical characteristics were found between Group A1 and A2.

Acute Safety of the Atrial Septum Biopsy

Serious complications directly associated with atrial biopsy were not observed, except for acute swelling of the atrial septum in 2 patients. ICE identified swelling of the atrial septum during the procedure immediately following the biopsy, which was subsequently confirmed by computed tomography after the ablation procedure. In these patients, pericardial effusion or conduction disturbances were not observed during and after the procedure. Ablation-related complications occurred in 3 cases (1 case of phrenic nerve paralysis and 2 cases of cardiac perforation), all of which were unrelated to the biopsy.

Long-Term Safety of the Atrial Septum Biopsy

Long-term follow-up imaging was performed using transthoracic echocardiography in 633 (69%) of 918 patients in the AF ablation group and in 31 (74%) of 42 patients in the non-AF ablation group at 1 year after biopsy. In the LAAO group, transesophageal echocardiography was performed in 30 (83%) of 36 patients at 4 months. In addition, 8 patients in the LAAO group underwent contrast-enhanced cardiac CT 1 year after the procedure. No endocardial defects, septal perforations, or hematomas were observed at the biopsy site in any of these follow-up assessments.

In ATTR amyloidosis cases identified by this atrial biopsy, tafamidis was prescribed for 7 patients, in accordance with the statement on the appropriate administration of tafamidis in patients with transthyretin cardiac amyloidosis issued by Japanese Circulation Society.¹⁵

Discussion

Major Findings

The main findings of this study are as follows: (1) ICE- or TEE-guided atrial biopsy is feasible and was successfully completed in 99.6% of patients without major complications, (2) only endocardial tissue was obtained in 14% of cases due to failure to penetrate the endocardium, and (3) risk factors for obtaining only endocardial tissue included larger atrial volume, smaller bioptome size, and the presence of amyloid deposition.

To our knowledge, this is the first study to report the feasibility of atrial biopsy at the right atrial septum in a large cohort of patients undergoing catheter ablation for atrial tachyarrhythmias or LAAO and identify risk factors for failure to obtain sufficient myocardial tissue for histological assessment. Despite being regarded as a high-risk procedure, this echocardiography-guided atrial biopsy technique has been shown to be safe.

Significance of Atrial Biopsy in the Diagnosis of Cardiac Amyloidosis

Recently, we reported an atrial amyloidosis prevalence of 7% in 578 patients with non-valvular AF undergoing catheter ablation.¹¹ Moreover, atrial amyloidosis was associated with the progression of atrial cardiomyopathy, characterized by reduced atrial voltage, and with a high risk of cardiovascular events.¹¹ Furthermore, concomitant right ventricular biopsy revealed ventricular amyloidosis in 50% of patients with atrial amyloidosis.¹¹ The present study builds on these findings by expanding the cohort to 1,000 patients, which included 42 patients without AF and 36 undergoing LAAO. The prevalence of atrial amyloidosis in the overall cohort remained consistent at 8%. Among these cases, 56% were classified as ATTR amyloidosis and 8% as AL amyloidosis for which early diagnosis and treatment are usually considered critical.^16–20

In patients diagnosed with cardiac amyloidosis through atrial biopsy, the average left ventricular wall thickness was 11±2 mm, indicating that left ventricular hypertrophy was not advanced.¹¹ However, in many of these cases, echocardiographic findings were not typical of cardiac amyloidosis. Furthermore, because technetium-99 m pyrophosphate scintigraphy was not performed prior to the biopsy, the scintigraphic positivity rate among patients diagnosed using atrial biopsy remains unknown. However, current imaging techniques, such as technetium-99 m pyrophosphate scintigraphy or computed tomography-based myocardial extracellular volume, may have insufficient sensitivity to detect ATTR cardiac amyloidosis early, often leading to diagnosis at a more advanced stage.^21,22 Consequently, it is difficult at present to preselect patients with a high likelihood of a positive biopsy result when performing atrial biopsy for the purpose of diagnosing cardiac amyloidosis. However, as shown in our recent study,¹¹ the probability of atrial amyloidosis can be estimated using parameters such as age and left ventricular wall thickness (±left atrial voltage), which may help in identifying appropriate candidates for atrial biopsy. Notably, most ATTR amyloidosis cases identified through atrial biopsy were diagnosed at an early stage and did not meet the current criteria outlined in the statement on the appropriate administration of tafamidis in patients with transthyretin cardiac amyloidosis issued by Japanese Circulation Society.¹⁵ As a result, tafamidis was prescribed in only 7 ATTR cases in this study. Therefore, the direct clinical benefit of this procedure remains limited at present. Further research is needed to identify appropriate candidates for atrial biopsy, clarify the role of early intervention in early-stage ATTR amyloidosis, and establish the clinical utility of atrial biopsy in this context. In addition, effective treatments for ANP-type atrial amyloidosis and confined AL-type atrial amyloidosis have not yet been established.^11,23 Future studies are warranted to develop therapeutic strategies for these conditions.

Potential Advantages of Atrial Biopsy Over Ventricular Biopsy

In our recent report, no cases of confined ventricular amyloidosis were observed, suggesting that atrial biopsy may potentially reduce the need for conventional right or left ventricular biopsy, which occasionally causes serious complications like cardiac perforation.^24,25 Moreover, right ventricular biopsy-associated right bundle branch block remains a major concern, particularly in patients with pre-existing left bundle branch block, because it may lead to complete atrioventricular block. Therefore, this echocardiography-guided atrial biopsy technique may serve as a safer alternative in such clinical settings.

Risk Factors for Failure in Obtaining Myocardial Tissue

Approximately 14% of all biopsy cases and 7% of cases using a 7.0-Fr bioptome resulted in the collection of only endocardial tissue without sufficient myocardial tissue. The identified risk factors included larger atrial volume, use of a 5.5-Fr bioptome, and amyloid deposition. A larger atrial size may be associated with a reduced residual myocardial layer or thickened and stiffened endocardium resulting from advanced atrial cardiomyopathy. In addition, atrial septal wall thinning secondary to atrial enlargement may have limited repeated biopsies at the same site because of the potential risk of perforation. Nevertheless, when using a 7.0-Fr bioptome, the atrial size was not identified as an independent predictor of failure to obtain sufficient myocardial tissue. The use of a 5.5-Fr bioptome may simply increase the risk of failure to obtain sufficient myocardial tissue due to the smaller sample size. No biopsy-associated complications were noted, even with the larger 7.0-Fr bioptome, indicating that the use of a larger bioptome may be more feasible for atrial biopsy. Last, amyloid deposition was identified as an independent risk factor in the overall cohort. Interestingly, it was the only significant risk factor in the subgroup of patients who underwent atrial biopsy using a 7.0-Fr bioptome. In all ATTR cases, amyloid deposition was consistently observed at least in the endocardium, which may have contributed to endocardial thickening and stiffening, resulting in the failure to penetrate the endocardium. Nonetheless, atrial biopsy remains a feasible and effective diagnostic approach for the detection of ATTR atrial amyloidosis as it can be diagnosed using endocardial tissue samples.

Sufficient Myocardial Tissue Group vs. Fat Layer Without Myocardial Tissue Group

Of the 885 patients in Group A, 867 (97%) had sufficient myocardial tissue for histological analysis (sufficient myocardial tissue group), whereas 18 (3%) had only endocardial and fat layer tissue (fat layer without myocardial tissue group). The presence of large amounts of fat tissue in the biopsy sample indicated that the bioptome had penetrated beyond the endocardium, and the procedure was considered technically successful. Interestingly, no significant differences were found in the clinical characteristics, bioptome size, or presence of amyloid deposition between the 2 groups. This finding is likely attributable to individual anatomical variations in fat layer distribution at the limbus of the FO.

Tips for Obtaining Sufficient Myocardial Tissue Samples

Based on our experience with 1,000 atrial biopsy cases, we propose the following tips: (1) select a target site at the limbus of the FO, where the operator can repeatedly position the bioptome at the same location, with clear visualization of the target site using ICE or TEE; (2) use a larger bioptome (7.0 Fr) to increase the likelihood of obtaining larger myocardial samples; (3) perform repeated biopsies at the same site – 3–4 biopsy attempts may be necessary in some cases to grossly obtain myocardial tissue (this technique is analogous to ‘digging a hole’); and (4) discontinue repeated biopsy attempts if the tissue samples float in the solution, as this suggests that the bioptome has reached the deeper fat layer, which is associated with a potential risk of perforation. These tips may facilitate the successful acquisition of sufficient atrial tissue without increasing the risk of complications, such as cardiac perforation. This atrial biopsy technique has potential applications in other clinical settings that use TEE, such as transcatheter aortic valve implantation and transcatheter mitral valve repair.

Study Limitations

This study has several limitations. First, atrial biopsies were performed by 5 experienced physicians at 4 institutions, which may limit the generalizability of the findings. Second, the possibility of sampling errors must always be considered. Third, in samples where only endocardial tissue was obtained, amyloid deposition in the myocardial layer may have been missed. Fourth, when atrial biopsy is performed for the purpose of diagnosing ATTR amyloidosis, the lack of technetium-99 m pyrophosphate scintigraphy in this study makes it difficult to determine the positivity rate. Therefore, it remains challenging to preselect patients with a high likelihood of diagnosis based on technetium-99 m pyrophosphate scintigraphy.

Conclusions

Endomyocardial atrial biopsy is a feasible and safe technique. Larger atrial size, smaller bioptome size, and the presence of amyloid deposition are associated with unsuccessful endocardial penetration and collection of sufficient myocardial tissue samples. This technique has potential clinical applicability, particularly for the diagnosis of atrial amyloidosis; although, the clinical significance of early diagnosis and therapeutic interventions remain unclear.

Disclosures

T.Y. received honoraria from Abbott Medical Japan and Medtronic Japan. T.O., T.Y., Y.T., and K. Nakashima were also affiliated with the Department of Advanced Management of Cardiac Arrhythmia, Saga University, sponsored by Abbott Medical Japan, Nihon Kohden Corporation, Japan Medtronic, Japan Lifeline, Boston Scientific Japan, and Fides-ONE Corporation. N.T. and K. Node are members of Circulation Reports’ Editorial Team. The other authors declare that they have no conflict of interest.

IRB Information

This study was approved by the Ethics Committee of Saga University Hospital (approval reference numbers: 20200101 for HEAL-AF, 20200901 for HEAL-AF2, 202200401 for FUTURE-AF-S, 20220401 for HEAL-AF3, 20240804 for HEAL-AF4, 20230203 for HISTORY, and 20230501 for LEARNMORE studies).

Supplementary Files

Supplementary File 1

Supplementary Table 1. Supplementary Table 2. Supplementary Figure.

Funding Statement

Sources of Funding: This work was supported by JSPS KAKENHI Scientific Research (A; 22H00471 to S.N., T.Y.) and Scientific Research (C; (JP21K08056 to T.Y., JP23K06434 to Y.S.-D.), Japan Agency for Medical Research and Development (JP22ek0210164, JP23ek0210164 to T.Y., K.I., K.Node; JP18 km0405209, JP23tm0724607, JP24ek0109755 to T.Y., S.N.).

Data Availability

Data are available upon reasonable request to the corresponding author.