Left Atrial Appendage Ostial Stenosis: A Case Series Authors: J Tyler Bruce, Tobias Z. Donlon, John L. Bass, Juan C. Samayoa Published: 2026 ABSTRACT Graphical abstract Introduction Primary stenosis of the left atrial appendage (LAA) ostium is a rare condition, with only 28 cases reported in adults and 2 in children.1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 Potential complications of this uncommon lesion are unknown; most reported cases were diagnosed incidentally and lack longitudinal follow-up. We report 2 infants with congenital membranous stenosis of the LAA and a third case of LAA ostial stenosis discovered after mitral valve replacement with a mechanical valve in the supra-annular position. Case Presentations Case 1 A 4-month-old male infant born at 35 weeks’ gestation with no underlying medical conditions underwent transthoracic echocardiography (TTE) as a workup for a murmur. The physical examination at this visit demonstrated a grade I-II/VI vibratory (innocent) systolic ejection murmur at the left lower sternal border with regular rate and rhythm and normal S1 and S2 and was otherwise unremarkable. TTE showed an abnormal color-flow Doppler signal of disturbed flow at the LAA entrance into the body of the left atrium (LA), with a membrane overlying the LAA ostium during atrial systole (Figure 1, Videos 1 and 2). The peak and mean gradients were 47 and 18.5 mm Hg, respectively (Figure 2).Figure 1Two-dimensional TTE, basal parasternal short-axis diastolic view without (A) and with (B) color-flow Doppler, demonstrates otherwise normal cardiac anatomy with a membrane originating at the LAA ostium (arrow), with flow acceleration noted at the LAA ostium (arrow). Ao, Aorta.Figure 2Two-dimensional TTE, subcostal long-axis view, demonstrates a spectral recording from continuous-wave Doppler from the LAA ostium. The peak velocity was 3.41 m/s, with a peak gradient of 47 mm Hg and a mean gradient of 18.5 mm Hg. Two-dimensional TTE, basal parasternal short-axis diastolic view without (A) and with (B) color-flow Doppler, demonstrates otherwise normal cardiac anatomy with a membrane originating at the LAA ostium (arrow), with flow acceleration noted at the LAA ostium (arrow). Ao, Aorta. Two-dimensional TTE, subcostal long-axis view, demonstrates a spectral recording from continuous-wave Doppler from the LAA ostium. The peak velocity was 3.41 m/s, with a peak gradient of 47 mm Hg and a mean gradient of 18.5 mm Hg. The remainder of the cardiac anatomy was normal, including normal size of the body of the LAA. On 6-month follow-up TTE, the color-flow Doppler signal was normal. The Doppler signal remained normal at 5-month follow-up. The child’s growth and development were normal, and the innocent heart murmur persisted at the last recorded visit. The patient never received anticoagulation or antiplatelet therapy during this course. Case 2 A 5-month-old female infant born at 24 weeks’ gestation underwent transthoracic echocardiographic screening for pulmonary hypertension in the setting of failure to wean from respiratory support at 7 months of age. Physical examination at the time of TTE showed a regular rate and rhythm, normal S1 and S2, and no associated murmurs. TTE in the apical 4-chamber view did not show any abnormality, but when color-flow Doppler was applied, an abnormal pattern was seen just inferior to the pulmonary vein going into the LA that also had retrograde flow into the LAA cavity during ventricular systole (Video 3, Figure 3). Peak and mean gradients of 26 and 5 mm Hg, respectively, were measured by continuous-wave Doppler from the LAA ostium (Figure 4).Figure 3Two-dimensional TTE, apical 4-chamber view without (A) and with (B) color-flow Doppler demonstrates normal cardiac anatomy with a membrane at the LAA ostium (arrow), with flow acceleration noted at the LAA ostium during diastole (B) and entering the LAA cavity during systole (C). MV, Mitral valve.Figure 4TTE, apical 4-chamber view, spectral tracing from continuous-wave Doppler, demonstrates the high-velocity flow with a biphasic signal. During atrial systole, a peak velocity of 2.53 m/s is recorded, with a peak gradient of 26 mm Hg and a mean gradient of 8 mm Hg. Two-dimensional TTE, apical 4-chamber view without (A) and with (B) color-flow Doppler demonstrates normal cardiac anatomy with a membrane at the LAA ostium (arrow), with flow acceleration noted at the LAA ostium during diastole (B) and entering the LAA cavity during systole (C). MV, Mitral valve. TTE, apical 4-chamber view, spectral tracing from continuous-wave Doppler, demonstrates the high-velocity flow with a biphasic signal. During atrial systole, a peak velocity of 2.53 m/s is recorded, with a peak gradient of 26 mm Hg and a mean gradient of 8 mm Hg. Seven previous screening transthoracic echocardiography examinations for pulmonary hypertension did not reveal the abnormal LAA signal, although imaging was complicated by high-frequency ventilation and patient agitation, and the region of the LAA ostium was not the focus of those studies. LAA ostial stenosis remained on multiple follow-up TTE through 12 months of age in the outpatient setting, with no associated symptoms or complications. The patient never received any anticoagulation or antiplatelet therapy. Case 3 A 13-year-old female patient with trisomy 21 underwent repair of a complete balanced atrioventricular (AV) canal defect at 5 months of age, followed by reoperation for left AV valve regurgitation 2 weeks later, with placement of a supra-annular 17-mm mechanical valve. The patient outgrew this valve, and it was surgically replaced at 10 years of age with a supra-annular 23-mm mechanical valve. Transesophageal echocardiography (TEE) performed 23 months postoperatively demonstrated an abnormal color-flow Doppler signal from the LAA ostium into the body of the LA and retrograde flow into the LAA cavity during ventricular systole (Figure 5, Videos 4 and 5). The flow out of the LAA split around a presumed suture of the mitral valve sewing ring. Peak and mean gradients of 64 and 28 mm Hg, respectively, were measured (Figure 6).Figure 5TEE, midesophageal LAA (47°) view, color-flow Doppler after mitral valve (MV) replacement, antegrade MV flow (blue signal) confirms diastolic timing and demonstrates that there is a jet of flow out of the LAA (A; red, arrow), and when the MV is closed there is a cleansing jet visible confirming ventricular systole with blue flow into the LAA (B; blue, arrow). Electrocardiogram not available.Figure 6TEE, midesophageal LAA (47°) view, continuous-wave Doppler across the LAA ostium obtained 23 months after mitral valve replacement, demonstrates abnormal flow (peak velocity 4.0 m/s, peak gradient 64 mm Hg, mean gradient 28 mm Hg). TEE, midesophageal LAA (47°) view, color-flow Doppler after mitral valve (MV) replacement, antegrade MV flow (blue signal) confirms diastolic timing and demonstrates that there is a jet of flow out of the LAA (A; red, arrow), and when the MV is closed there is a cleansing jet visible confirming ventricular systole with blue flow into the LAA (B; blue, arrow). Electrocardiogram not available. TEE, midesophageal LAA (47°) view, continuous-wave Doppler across the LAA ostium obtained 23 months after mitral valve replacement, demonstrates abnormal flow (peak velocity 4.0 m/s, peak gradient 64 mm Hg, mean gradient 28 mm Hg). Follow-up TTE did not show a significant change in gradient or intracavitary atrial appendage appearance over the course of 6-year follow-up. The patient remained clinically asymptomatic from the LAA ostial stenosis. They remained on anticoagulation for their mechanical valve. Discussion Ostial stenosis of the LAA is an extremely rare and poorly characterized entity in both adult and pediatric patients. The medical literature primarily describes isolated case reports and small series, with no consensus regarding the clinical significance, natural history, or management. There are at least 30 cases described in the literature, all but 2 in adults. Ages of the 28 adult patients ranged from 26 to 84 years (average, 62.3 years). Seventeen of the adult patients were in atrial flutter or fibrillation.3, 4, 5, 6,9,10,12,14,15 Smoke was observed in the LAA in only 1 case.4 To date, no pediatric patients have been reportedly anticoagulated, reflecting the presumption that thromboembolic risk in this population is low in the absence of atrial arrhythmia or other predisposing conditions. In our cohort, the providers did not prophylactically anticoagulate these patients given unclear benefit. Associated diagnoses of rheumatic fever (1 case)7 and asymmetric septal hypertrophy (1 case)12 were uncommon. It seems likely that these associated conditions of arrhythmia or structural heart disease were indications for TEE to be performed, uncovering the diagnosis, rather than a primary cause of arrythmia. In a report of 6,000 transesophageal echocardiographic studies in patients with atrial fibrillation reviewed retrospectively, only 6 cases of LAA ostial stenosis were found.9 Furthermore, with the interest in device exclusion of the LAA to reduce stroke risk, extensive reviews of LAA anatomy were conducted with TEE and pathology series.18, 19, 20, 21, 22 These reviews failed to identify any cases of this membrane of the LAA ostium. Two forms of obstruction have been described: a tunnel-like ostial stenosis (5 cases)4,5,13,16 or a membrane partially covering the LAA ostium (8 cases).1,2,7,9,10,12,15,17 There are multiple reports of a membrane’s being present at the LAA ostium but without obstruction, including a pediatric patient.3,6,8,11,14,23 Most authors favored a congenital etiology, as it is difficult to explain how an atrial arrhythmia would lead to the formation of a membrane. It is perhaps unexpected that the LAA can generate such high velocities. Peak velocities in atrial systole in the 6 reported cases with sinus rhythm ranged from 1.6 to 2.9 m/s (Table 1) and from 2.6 to 3.2 m/s in our congenital pediatric cases, and peak velocity was 4.3 m/s in our third case.Table 1Clinical features in reported cases of obstructive LAA ostial stenosisAuthorsAge, yr/genderRhythmVelocity, m/sMembrane or stenosisThrombus presentMcCartney et al.7 (2017)47/maleNA1.4MNoJonas et al.14 (2019)70/maleAFNAMNoCresti et al.9 (2022)72/maleAF or AFLNAMNoKim et al.5 (2010)70/femaleAF1SNoHanke et al.10 (2023)56/maleAF1MNoKim et al.5 (2010)68/maleAFL1.1SNoKim et al.12 (2017)55/femaleAFL1.7MNoLi et al.15 (2022)48/maleAFL1.5MNoCoughlan et al.1 (1999)66/femaleSR1.6MNoKim et al.5 (2010)66/femaleSR1.6MNoStern et al.4 (2009)83/maleSR1.6SNoHa et al.2 (2001)45/femaleSR1.8MNoMu et al.13 (2019)46/maleSR2SNoGarcía et al.17 (2023)0.33SR2MNoOkada et al.16 (2016)0.25/maleSR2.9SNoBruce et al. (2026)0.42/femaleSR2.5MNo0.33/femaleSR3.4MNo15/femaleAtrial paced3.9SNoAF, Atrial fibrillation; AFL, atrial flutter; F, female; M, membrane; NA, not available; S, stenosis; SR, sinus rhythm. Table 1Clinical features in reported cases of obstructive LAA ostial stenosisAuthorsAge, yr/genderRhythmVelocity, m/sMembrane or stenosisThrombus presentMcCartney et al.7 (2017)47/maleNA1.4MNoJonas et al.14 (2019)70/maleAFNAMNoCresti et al.9 (2022)72/maleAF or AFLNAMNoKim et al.5 (2010)70/femaleAF1SNoHanke et al.10 (2023)56/maleAF1MNoKim et al.5 (2010)68/maleAFL1.1SNoKim et al.12 (2017)55/femaleAFL1.7MNoLi et al.15 (2022)48/maleAFL1.5MNoCoughlan et al.1 (1999)66/femaleSR1.6MNoKim et al.5 (2010)66/femaleSR1.6MNoStern et al.4 (2009)83/maleSR1.6SNoHa et al.2 (2001)45/femaleSR1.8MNoMu et al.13 (2019)46/maleSR2SNoGarcía et al.17 (2023)0.33SR2MNoOkada et al.16 (2016)0.25/maleSR2.9SNoBruce et al. (2026)0.42/femaleSR2.5MNo0.33/femaleSR3.4MNo15/femaleAtrial paced3.9SNoAF, Atrial fibrillation; AFL, atrial flutter; F, female; M, membrane; NA, not available; S, stenosis; SR, sinus rhythm. Clinical features in reported cases of obstructive LAA ostial stenosis AF, Atrial fibrillation; AFL, atrial flutter; F, female; M, membrane; NA, not available; S, stenosis; SR, sinus rhythm. Reported lower velocities were found in patients with atrial arrhythmias (atrial fibrillation or flutter). Lower velocities with an LAA membrane may be a result of disorganized atrial contraction. It was shown that the LAA can generate high pressures in sinus rhythm. The continuous-wave Doppler velocity profile seen with obstructive LAA ostial membrane is biphasic, with a high-velocity signal during atrial systole, preceded by a lower velocity signal during “passive” ventricular filling (Figure 2, Figure 4, and 6). This is also true in published cases in sinus rhythm.1,2 The LAA membrane and abnormal color-flow Doppler pattern were best seen from the parasternal short-axis and subcostal windows (Figures 1 and 3). Although the abnormal color-flow Doppler pattern could be identified from other windows, it was less obvious. The subcostal window was most parallel to flow on TTE in our series and yielded the highest velocity measurements. Our report represents the third and fourth pediatric cases of membranous ostial stenosis, including a case of spontaneous resolution. We also describe LAA stenosis occurring after left AV valve replacement with a mechanical valve. Previously reported cases have rarely been associated with other anomalies such as prematurity, cor triatriatum, or prior surgical manipulation of the LA.16,17,24 Two of our patients were premature infants with absence of additional risk factors suggesting a congenital etiology. The first patient in this case series presented with a systolic ejection murmur considered to be an innocent murmur. In our third case, their age and the presence of prior cardiac surgery complicated attribution of cause but suggest an acquired etiology. In addition to the two reported cases by Okada et al.16 and García et al.,17 we present two new cases of LAA stenosis in children born prematurely, suggesting that prematurity of birth may increase the risk for LAA stenosis. Conclusion LAA ostial stenosis is an extremely rare defect in both pediatric and adult patients. When LAA ostial stenosis is present, careful evaluation is needed to guide care. Longer term follow-up of these cases is warranted to better understand the long-term prognosis of this rare lesion. Ethics Statement The authors declare that the work described has been carried out in accordance with the following guidelines: The study was conducted in accordance with the Declaration of Helsinki. Ethical approval was not required for this case report, as it involved routine clinical care and no experimental interventions. Consent Statement The authors declare that since this was a non-interventional, retrospective, observational study utilizing de-identified data, informed consent was not required from the patient under an IRB exemption status. Disclosure Statement Dr. Samayoa is a consultant and proctor for B. Braun.