Intussusception of the Left Atrial Appendage: A Systematic Review

Abstract

Background

Intussusception of the left atrial appendage (LAA) is an uncommon and often underrecognized finding; poor knowledge of this condition is still widespread in clinical practice. A systematic review was carried out to provide a comprehensive and updated overview on this condition.

Methods

A bibliographic search using major databases resulted in 25 case reports meeting the criteria. Included records reported data from patients aged > 18 years with a diagnosis of LAA intussusception, determined either intraoperatively or incidentally by imaging tools, and findings on baseline demographics, clinical findings, imaging diagnostic workup, and management.

Results

For the 25 patients included, the median age was 54 years, with a predominance of female sex (53%). Most of intussuscepted LAA cases were disclosed intraoperatively (60%) or incidentally by imaging tools (20%) in the absence of associated clinical findings. All reported cases underwent transoesophageal echocardiography, and in most, LAA intussusception was confirmed by intraoperative inspection (44%) or further imaging (23%). In most cases, LAA intussusception was resolved interventionally (60%). In fewer subjects, LAA eversion occurred spontaneously (32%), and in the remaining cases, a conservative approach was adopted (8%).

Conclusions

This study presents the largest systematic review to date of an adult patient population with a diagnosis of intussuscepted LAA. Regardless of the various etiologies and clinical manifestations associated with LAA intussusception, physicians should be aware of the possibility of this rare entity, to provide timely recognition and thereby avoid misleading diagnoses and enable appropriate decision-making.

Intussusception of the left atrial appendage (LAA) is an uncommon and underrecognized finding in clinical practice. The condition often develops as a procedural sequela of interventional heart procedures, and more rarely, it occurs spontaneously, in subjects without a history of cardiac interventions. Clinical manifestations related to an intussuscepted LAA are heterogeneous and often unpredictable. They range from its incidental appearance on imaging in asymptomatic subjects to dangerous and potentially life-threatening complications, such as mitral inflow obstruction, reduced stroke volume, pulmonary hypertensive crisis, compromised appendicular wall perfusion, and mechanical appendicular damage. Proper identification of LAA intussusception by imaging tools, and its differential diagnosis vs other left atrial masses, is often challenging. Despite sporadic cases and case series reported in the literature, poor knowledge of this infrequent entity and its anatomic features is still widespread in clinical practice.^1,2 This review aims to provide a comprehensive and updated overview focusing on the etiopathogenesis, clinical findings, imaging diagnostic workup, and therapeutic approaches for LAA intussusception, to increase knowledge of this rare entity and its potential implications in clinical practice.

Methods

A systematic review was carried out by following the standard methods of the Cochrane Collaboration and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. The keywords “left atrial appendage” AND (“intussuscepted” OR “intussusception” OR “invaginated” OR “invagination” or “inverted” OR “inversion”) were searched in PubMed, MEDLINE, Embase, and the Cochrane Central databases, up to December 31, 2024. Two reviewers (R.S. and G.P.) independently and separately collected data using a standardized data abstraction form. A third reviewer (M.B.) adjudicated any discrepancies during the screening process (Fig. 1).³ Of the 115 potential records initially retrieved for inclusion in the systematic review, 27 duplicates and 2 records in languages other than English were removed. Following a thorough screening of the literature, 32 records were ruled out, as they did not pertain to the research topic. Other excluded items were the following: papers without full-text available; original articles; conference proceedings; editorials; letters to the editor; and author’s replies. Records also were excluded if they were pediatric data, lacked individual case data, or had insufficient data regarding baseline demographics, clinical findings, imaging diagnostic workup, and management. Overall, 25 records were identified for final inclusion (Supplemental Table S1). Extracted data were pooled for analysis and reporting. Continuous variables were expressed as median values, and categorical variables were described as proportions (%). Statistical analysis was carried out using the Statistica 13.1 software for Windows (TIBCO StatSoft, Tulsa, OK).

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) diagram depicting the literature selection process and included records.

Results

A total of 25 cases were included. Patient demographics included a median age of 55 years (interquartile range [IQR]: 37.5-61.5), and a predominance of female cases (n = 13; 53%). Most of identified cases came from the US (n = 9; 36%). The majority of subjects developed LAA intussusception following interventional heart procedures, either surgical or percutaneous (n = 20; 80%). Atrial fibrillation and coronary artery disease were the most frequently associated comorbidities, with equal prevalence among the included cohort (n = 4; 16%). Most of the intussuscepted LAA cases were discovered intraoperatively (n = 15; 60%), or incidentally by imaging tools (n = 5; 20%), in the absence of associated clinical findings. All reported cases underwent transoesophageal echocardiography (TOE), either intraoperatively or in a noninterventional clinical context. However, in most of the cases, LAA intussusception was confirmed directly by intraoperative inspection (n = 11; 44%), or alternatively by further imaging, such as cardiac magnetic resonance imaging (MRI) or computed tomography (n = 5; 23%). In most cases, intussuscepted LAA resolved interventionally (n = 15; 60%), whereas in fewer subjects, LAA eversion occurred spontaneously, following heart blood flow restoration (n = 8; 32%), and in the remaining cases, a conservative approach was adopted (n = 2; 8%; Table 1).

Table 1.

Baseline findings in included cohort

Demographics	Value
Gender, female	13 (53)
Age, y	55 (37.5–61.5)
Country
US	9 (36)
Europe	8 (32)
Asia	5 (20)
Other	3 (12)
Comorbidities
Atrial fibrillation	4 (16)
Coronary artery disease	4 (16)
Arterial hypertension	3 (12)
Hyperlipidemia	1 (4)
Associated clinical findings
None	20 (80)
Heart failure	2 (8)
Cardiogenic shock	1 (4)
Cardiac tamponade	1 (4)
Ischemic stroke	1 (4)
Imaging tools
Transoesophageal echocardiography	25 (100)
Cardiac magnetic resonance imaging	3 (12)
Cardiac computed tomography	2 (8)
Management
Interventional reduction	15 (60)
Spontaneous eversion	8 (32)
Conservative approach	2 (8)

Values are n (%) or median (interquartile range).

Discussion

Definition and historical background

Intussusception of the LAA (also reported as LAA inversion or invagination) is defined as a complete or partial auricular introflexion into the left atrial chamber.⁴ Historically, the first description of LAA intussusception was provided in 1979, by Roberts et al., as a gross autoptic finding at the necropsy of a 27-year-old man who died after a gunshot wound, in whom LAA was found to be introflexed into the mitral valve orifice.⁵ In 1992, Aronson and colleagues reported the first echocardiographic appearance of intussuscepted LAA as a newly developed left atrial mass, immediately prior to the discontinuation of cardiopulmonary bypass (CPB) after a surgical intervention of left ventricular thrombectomy.⁶ Subsequently, in 1997, Toma and coworkers provided further echocardiographic findings for characterizing the intussusception of LAA, and further imaging clues to use for differential diagnosis of this uncommon entity vs other potentially confounding left atrial masses.⁷

Etiologies

In our analysis, LAA intussusception was noticed as an iatrogenic finding following cardiac interventions, related to either cardiac surgery or percutaneous heart procedures. Invaginated LAAs have been detected mostly intraoperatively during open-heart surgerythat required CPB to temporarily provide a bloodless field for the surgeons.^2,8 Intussuscepted LAA also has been described following minimally invasive cardiac surgery, and as an iatrogenic complication of percutaneous structural heart interventions.9, 10, 11 Intussusception of the LAA also was reported anecdotally to have occurred spontaneously, in subjects without history of cardiac intervention, and in cases in which the etiopathogenetic mechanism was not related directly to interventional heart procedures.¹² Cases of idiopathic LAA intussusception in subjects without likely triggers are extremely rare, and de novo LAA intussusceptions have been noticed in only a few cases in the presence of a predisposing condition (eg, paroxysmal supraventricular arrhythmias, volume loss, left-to-right shunts, or increased intrapericardial pressure).13, 14, 15, 16, 17

Pathogenic mechanisms and predisposing morphologic features

Several pathogenic mechanisms have been proposed to explain the intussusception of the LAA in various clinical settings. One of the commonest postoperative conditions predisposing patients to LAA intussusception is the excessive negative pressure that occurs through the left atrial vent catheter during weaning from CPB, while the heart is still empty.^2,18 A similar contributing factor is the excessive negative pressure created by de-airing maneuvers performed before discontinuation of CPB (these include digital air evacuation performed intraoperatively by the surgeon’s manipulation of the heart).¹⁹ In the surgical setting, LAA intussusception also can be caused by its external compression due to continuous expansion of the left lung by the delivery of high continuous positive airway pressure. This possibility is especially likely in minimally invasive mitral valve surgery, in which this technique usually is applied to push the heart in a rightward direction, thereby increasing the exposure of the mitral valve and squeezing the LAA.^9,20 Of note, iatrogenic factors that could be responsible for complete or partial intussusception of the LAA also have been reported following percutaneous heart procedures. These factors are related to either valvular interventions or intracardiac device implantations, particularly in cases of disproportionate sheath advancement and traction on the device.^9,11 A decrease in ventricular volume and a shortening of the diastolic time of the cardiac cycle also have been described as likely mechanisms for creating a predisposition to LAA intussusception. This mechanism is theoretically applicable in several circumstances. Goldreyer et al. showed the hemodynamic effects of fast ventricular emptying and reduced refilling triggered by supraventricular tachycardia.²¹ Kato and colleagues elucidated the mechanisms of LAA stunning as a consequence of electromechanical dissociation of the LAA during paroxysmal atrial fibrillation. This electrical behaviour could have an impact on the physiological mechanics of the human heart, creating transitory negative pressures within the left atrium, thereby producing a vacuum effect during the early diastole, which is capable of invaginating the LAA.^15,22 Analogous speculations could be applied to explain other likely conditions of volume loss. In the first reported case by Roberts and coworkers, intussusception of the LAA developed because of an extreme decrease of left ventricular preload due to traumatic right-sided exsanguination, which created negative pressure and a suction effect in the left-sided heart chambers.⁵ Similarly, a negative pressure gradient also may result from a disproportional increase in intrapericardial pressure (as in the case of massive effusions), or secondary to left-to-right intracardiac shunts, thus favouring auricular introflexion into the left atrium.^17,23,24 In addition, the risk of LAA intussusception is influenced strongly by the structural characteristics of the LAA. Longer and thinner appendages with a narrow orifice (as in the case of “chicken-wing” morphology) have been shown to be prone to a higher risk of sustained LAA intussusceptions. They are also susceptible to greater difficulty in spontaneously recovering their physiological configurations, as they require the lowest suction pressure for achieving intussusception into the left atrium, and the highest pressure to restore its normal position, as compared to other LAA shapes. Additionally, they are linked to a higher risk of potential interference with adjacent cardiac structures, such as the left-sided pulmonary vein orifices or mitral valve inlet, due to a longer and angulated protrusion into the left atrial chamber. Conversely, shorter appendages with a high number of trabeculations and a broad orifice (such as that for the “cactus” shape) have been shown to be particularly safe, due to their tendency to invert either partially or in a more compact configuration. For these reasons, they are associated with a lower risk of interference with contiguous structures, as compared to other LAA morphologies.^17,25

Clinical findings

Regardless of its causative agents, be they iatrogenic or not, intussusception of the LAA may lead to a wide range of clinical manifestations; it is occasionally detected in asymptomatic subjects, and it sometimes creates deranging and potentially life-threatening complications. In our analysis, most cases of LAA intussusception were noticed intraoperatively or incidentally by imaging tools, in patients without associated clinical findings. More rarely, LAA intussusception was detected in symptomatic subjects. The most dangerous hemodynamic sequela relate to the overall effect caused by an intussuscepted LAA is its introflexion into the mitral valve orifice. The mitral inflow obstruction caused by such introflexion leads to impaired ventricular filling pressures, reduced stroke volume, pulmonary hypertensive crisis, and associated acute right ventricular failure. This state, in turn, poses a clinical challenge within this critical context, as the first-line therapy for right-sided acute overload requires diuretic treatment, which may exacerbate the hemodynamic impairment if invaginated LAA is not suspected and properly managed within the appropriate timeframe.^8,26 Other possible complications of LAA intussusception include hemorrhagic transformation, adhesions, and incarcerations with impending appendicular necrosis from a compromised perfusion of the appendicular wall.^2,27 On the other hand, its potential thrombogenic role remains unclear. A single case of ischemic stroke with a possible embolic cause has been reported by Ankersmit et al. in a patient with contextual detection of de novo intussusception of the LAA, suggesting that it may have thrombogenic properties.²⁸ Conversely, Allen and colleagues presume LAA intussusception to be a benign natural occurrence that should not induce clot formation, as it is completely endothelialized, lies directly on the path of rapid blood flow, and no longer has a shape favouring blood stasis.²⁹ Additional data are needed to assess the relationship between LAA intussusception and thromboembolic complications.

Imaging tools and differential diagnosis

The suspicion of LAA intussusception poses several questions relating to the differential diagnosis of several kinds of left atrial masses; the inclusion and and exclusion of sources should be performed based on imaging³⁰ (Table 2). TOE is the imaging modality of choice for detecting intra-atrial masses, and it is a useful tool for both confirming the suspicion of invaginated LAA, in the appropriate clinical context, and ruling out other confounding lesions, due to its optimal visualization of the left atrium and the LAA, which lies in the near field of the probe⁴ (Fig. 2). Specific echocardiographic features of suspected LAA intussusception include the following: (i) a newly appearing mass in the left atrium; (ii) such a mass with smooth margins and fairly homogeneous density, and sonographic tissue texture analogous to that of the atrial wall; (iii) projection of such a mass from the anterolateral left atrial wall, between the ostium of the left upper pulmonary vein and the annular attachment of the posterior mitral leaflet, toward the mitral valve or intermittently flopping into the mitral orifice in diastole; (iv) such a mass without attachments to the mitral valve or the atrial septum; (v) lack of flow signal through the mass upon color-flow Doppler interrogation; (vi) simultaneous failure in visualizing LAA, and the abrupt resolution of the mass with subsequent LAA visualization, once left atrial pressure and volume have been normalized.^6,24,31,32

Table 2.

Characteristic clues to use to identify left atrial masses

Atrial mass	Characteristics
LAA intussusception	-Newly apparent left atrial mass -Sonographic tissue texture analogous to that of atrial wall -High mobility -Between left-sided pulmonary veins ostia and mitral annulus -Lack of flow signal at colour-flow Doppler interrogation -Lack of attachments to the mitral valve or atrial septum -Failure in visualizing LAA
Atrial myxoma	-Septal location, at the region of the fossa ovalis -Anchored by a narrow pedicle -Lack of LAA involvement -Lack of laminated appearance -Rough core with fairly heterogenous density
Atrial thrombus	-LAA location -Laminated appearance -Different echo density from that of the left atrial wall -Irregular or lobulated borders -Preexisting thrombogenic conditions
Infective endocarditis	-Involvement of valve apparatus and/or intracardiac devices -Erratic motion -Clinical suspicion of infective endocarditis
Prominent left atrial ridge	-Located between the LAA and left superior pulmonary vein -Low mobility -Contextual identification of LAA

LAA, left atrial appendage.

Figure 2.

(A) Mid-esophageal bidimensional transoesophageal echocardiography 4-chamber view showing a large left atrial mass (arrow). (B) Mid-esophageal bidimensional transoesophageal echocardiography 2-chamber view revealing the mobile mass arising from the anterolateral wall of the left atrium, between the ostium of the left upper pulmonary vein and the mitral annulus (arrow; reproduced from Nunes et al. J Echocardiogr 2010;8:30-2, with permission of the publisher).¹⁵ LA, left atrium; LUPV, left upper pulmonary vein; LV, left ventricle; MA, mitral annulus; RA, right atrium; RV, right ventricle.

Because of its high level of mobility, the imaging plane that cuts the invaginated LAA may show different morphologic shapes along different axes at different times in the cardiac cycle. When intussuscepted LAA is cut longitudinally, it will at times have the appearance of a slender, pyramidal structure, giving the appearance of a “crooked finger” or “‘tongue,” both of which are highly suggestive of its identity. Alternatively, when intussuscepted LAA is cut cross-sectionally, it will have an irregular, somewhat globular appearance, which may raise doubt about its identity.^13,16,27,33

The main features on which to base a differential diagnosis of a suspected LAA intussusception include intra-atrial tumours, infective vegetations, and thrombi. A mass that newly develops intraoperatively or in the immediate postoperative period excludes the first 2 of these features. This approach requires that a preoperative TOE has been obtained.^1,22,34 LAA intussusception may be mistaken for a thrombus in the postoperative setting. Typical identifying features for a left atrial thrombus include the following: (i) its presence in the LAA; (ii) a laminated appearance; (iii) an echo density that usually differs from that of the left atrial wall; and (iv) irregular or lobulated borders with associated spontaneous microcavitations. Additionally, the likelihood of a left atrial thrombus is low in the presence of the following: preexisting preventing conditions, such as sinus rhythm; the assumption of therapeutic anticoagulation; and/or the absence of moderate-to-severe mitral valve stenosis.^6,8

Another diagnosis that may be difficult to differentiate from LAA intussusception includes a prominent and well-developed left atrial ridge—that is, a fold of the left lateral atrial wall containing the ligament of Marshall (a remnant of the embryonic sinus venosus and left cardinal vein). However, its unique location between the LAA and left-sided pulmonary veins ostia and its low mobility are useful clues for differentiating this from LAA intussusception.^35,36 Another benign finding that sometimes occurs, and should not be mistaken for LAA intussesception for patients underoing cardiac surgery is the echocardiographic appearance of the surgeon’s finger.^7,37 Other imaging modalities are seldom used for investigating left atrial masses when echocardiographic responses are unclear. Cardiac MRI is sometimes used to address doubts concerning the correct etiology of intra-atrial masses. Cardiac MRI also is used to better characterize LAA intussusception with further imaging findings, including the following: (i) a high attenuation of the signal intensity, due to presence of fat tissue within the lesion on fat-suppressed black-blood fast spin echo MRI (which indicates the presence of fat tissue within the lesion in broad continuation to the epicardial fat); and (ii) the cyclic protrusion of this fat-containing structure inward to the left atrial lumen, which is especially apparent during the end-diastolic phase on steady-state cine MRI (Fig. 3).^13,17,38,39 However, findings indicate that this imaging modality is not diriment in all cases, and to date, the appropriateness of its standard routine use in this clinical setting is still debatable.²⁸ Electrocardiogram-gated cardiac computed tomography theoretically could be a valid complimentary imaging tool, despite the lack of properly established confirming radiologic criteria. Nevertheless, several imaging clues, such as evidence of a low-density mass-like lesion arising from the anterolateral aspect of the left atrium, and the demonstration of a geometric relationship with the left upper pulmonary vein using a 3-dimensional reconstruction may be helpful in reaching the correct diagnosis (Fig. 4).^8,40,41 If the above-mentioned imaging modalities do not yet clarify suspicion of an intussuscepted LAA, a direct atrial visualization from the surgical field is sometimes required to confirm the diagnosis.

Figure 3.

Cardiac magnetic resonance imaging showing (A, B, thin arrows) the intussusception of left atrial appendage mimicking a cystic mass in the left atrium (LA);and (C, D, thick arrows) left atrial appendage being pushed and invaginated into the LA by a massive pericardial effusion (PE) ; reproduced from Gecmen et al.¹⁷ Echocardiography 2011;28:E134-6, with permission of the publisher).

Figure 4.

(A) Transoesophageal echocardiography showing a tongue-like mass (arrows) with a central high-echogenic portion protruding to the left atrium (LA). (B, C) Cardiac computed tomography (CT) showing a mass-like lesion with internal fat tissue (asterisks) and an air bubble (arrowhead) arising from the anterolateral wall of the LA between the left upper pulmonary vein (LUPV) and the mechanical mitral valve (curved arrow). (D) A 3-dimensional endoluminal of the CT image obtained during the diastolic phase revealing the intussuscepted left atrial appendage (arrows) at the posterior aspect of the mitral inflow tract, with no significant obstruction. (E) A follow-up CT scan obtained at 15 months showing the spontaneous eversion of the left atrial appendage (arrows; reproduced from Yoon et al. Circulation 2014;130:e66-7, with permission of the publisher).⁴¹ Ao, ascending aorta; LCx, left circumferential artery; LV, left ventricle; PA, pulmonary artery.

Natural history and treatment

To date, the optimal management for LAA intussusception is debatable. Surgical correction and a conservative approach are both plausible options. So far, no approach has been proven to be superior to the others. Although LAA intussusception may be transient and reversible, its spontaneous eversion cannot be predicted easily, as long-term follow-up investigations are lacking.^1,38 Intussusception of the LAA often reverts spontaneously once the heart is refilled by blood volume normalization, and left atrial pressure gradually increasing up to normal values.^23,24,27 In addition, cases of spontaneous reversal of non-iatrogenic LAA intussusceptions have been reported once the underlying pathogenic triggers have resolved.^12,14 In the case of unsuccessfully spontaneous eversion, the Valsalva maneuver may be applied to transiently increase left atrial pressure markedly beyond its normal levels. However, if this attempt had proven unsuccessful, a careful assessment of the patient’s clinical and hemodynamic profile is mandatory.²⁰ Given that even irreversible LAA intussusceptions can be asymptomatic, a conservative approach with close and careful follow-up may be considered in this subset population, given the risk associated with reinterventions. Conversely, for patients in whom the persistence of intussuscepted LAA leads to an unstable hemodynamic profile, a further interventional approach is warranted for preventing rare but potentially life-threatening complications.^14,23 Digital manipulation or pulling with forceps are the usual surgical modes of LAA eversion, and catheter-based interventional LAA repositioning also has been reported.^35,42 Ligation or resection of the LAA is generally unnecessary, unless peri-procedural tissue injury has occurred. No evidence has shown that LAA ligation prevents its intussusception over time, and cases of reinverted LAA following ligation also have been reported.^27,34,43 Therefore, the therapeutic approach for management of LAA intussusception should focus on the patient’s hemodynamic features, to maximize therapeutic benefit by minimizing the risk of inappropriate surgical or percutaneous reinterventions. Despite the fact that systemic anticoagulation often is pursued as a preventive in these kinds of patients, no data support use of this pharmacologic approach, as to date, no causative relationship between the occurrence of invaginated LAA and thromboembolic events has been determined.^4,17,29

Limitations

This systematic review has several limitations. First, the included cohort was drawn exclusively from single-case data, thus resulting in publication bias. In addition, the collected data are limited by the small sample size and discrepancies among cases in the quantity and quality of the data reported. The analysis also lacks a control group, because of the paucity of available data. Despite the aforementioned limitations, this systematic review is the largest to date in a single set of adult patients with a diagnosis of intussuscepted LAA, and with complete findings on epidemiology, etiologies, clinical presentation, diagnostic workup, and management. Further investigation in this setting is needed, to provide a larger and more representative clinical dataset.

Conclusions

Intussusception of the LAA remains an uncommon and often underrecognized finding in clinical practice. The condition has various etiologies—either iatrogenic or unrelated to cardiac interventions—and a heterogeneous set of associated clinical manifestations. Physicians should be aware of this rare possibility, so that timely recognition occurs, either via imaging tools or direct intraoperative inspection, thus avoiding misleading diagnoses and enabling appropriate decision-making.