Anatomy of the Normal Left Atrial Appendage
A Quantitative Study of Age-Related Changes in 500 Autopsy Hearts: Implications for Echocardiographic Examination
Background Transesophageal echocardiography (TEE) is the diagnostic modality of choice for visualizing the left atrial appendage (LAA). This study defined the morphology of the LAA in normal autopsy specimen hearts and considered the implications of these findings for TEE studies.
Methods and Results Five hundred normal autopsy hearts were reviewed (25 male and 25 female subjects from each decade for 10 decades). LAA length, width, orifice size, and number of lobes were recorded. Number of lobes was compared between sexes with the rank sum test and regressed against age. Mean length, width, and orifice size increased with age, up to age 20 years, in both sexes. Rates were significantly different between sexes for LAA size (P=.011) and width (P=.006). After age 20, statistically significant but clinically insignificant age-related changes were observed. Fifty-four percent of LAAs had two lobes (range, 1 to 4), with no age or sex differences. Lobes exist in different planes of the heart. Most pectinate muscles were ≥1 mm in width. Pectinate muscles <1 mm (2.6% of cases) were seen in only the first and last decades.
Conclusions Age- and sex-related differences in LAA dimensions exist. These differences and the existence of multilobed appendages are important in the accurate TEE evaluation of LAA. Because lobes exist in different planes, imaging must be done in multiple planes to visualize the entire LAA.
Transesophageal echocardiography (TEE) is an integral part of any comprehensive ultrasonographic examination of the heart and great vessels because the proximity of the esophagus to the heart and great vessels permits high-quality imaging in nearly all patients undergoing TEE.1 This technique is safe to perform, and the list of indications continues to increase.1 Emboli from an intracardiac source are an important and frequent cause of transient ischemic attacks and cerebrovascular accidents. The left atrial appendage (LAA) is often suspected of harboring the source of embolic material. Recent investigations1 2 3 4 5 6 7 8 have focused on the clinical usefulness of TEE and its superiority over transthoracic echocardiography in the evaluation of the LAA as a source of thromboembolism; in fact, this is one of the most common indications for performing TEE.1 Recent studies have proposed the use of TEE before cardioversion for detection of left atrial and LAA thrombus in patients with atrial fibrillation.9 10 This recommendation is based on a commonly entertained but unproved theory that the absence of thrombus in the left atrium and LAA, as determined with TEE, may safely allow early cardioversion without prior prolonged anticoagulation.
With the increasing use of TEE for morphological and functional evaluation of the LAA, it has become important to obtain information about normal LAA anatomy. Because we could not find a detailed study of the normal LAA, this study was initiated.
Specimens of 500 normal human hearts from autopsies performed between 1960 and 1982 were procured from the Mayo Clinic tissue registry.11 12 13 14 Specimens were excluded if a clinical history of cardiac disease or pulmonary or systemic hypertension was present or if any evidence of heart disease was noted at autopsy. All specimens were reexamined grossly to ensure exclusion of abnormal specimens. Specimens with nonintact or nonreconstructable LAA were excluded from the study.
We obtained specimens from 25 male and 25 female subjects from each decade of life, ages 1 to 100 years.
For each heart, measurements of LAA length, width, orifice size, and number of lobes were recorded (Figs 1⇓ and 2⇓). The LAA shape and presence or absence of pectinate muscles and their size were also recorded.
The left atrium was opened and the orifice size was measured, with a ruler calibrated in centimeters, from a point just inferior to the left superior pulmonary vein to the opposite side of the orifice opening (Fig 1⇑). The shape was elliptical rather than round. The orifice width was taken as the maximum dimension of the ellipse. Length was measured as the distance perpendicular from the opening of the orifice to the apex of the LAA tip. The widest external width was noted.
The number of lobes was counted by external examination and confirmed by probe exploration after the LAA was opened. Pectinate muscle size (<1 mm or ≥1 mm) was recorded.
A lobe was defined by the following criteria: (1) it was a visible outpouching from the main tubular body of the LAA, usually demarcated by an external crease; (2) it was internally capable of admitting a 2-mm probe (ie, it was not simply a tag of external adipose tissue); (3) it was occasionally but not necessarily associated with a change in direction of the main tubular body of the LAA; (4) it could lie in a different anatomic plane than the main tubular body; and (5) by definition, the LAA must have at least one lobe (ie, a tubular body with a blind-ending sac).
The distribution of each measurement was examined overall and by age category (<20 or ≥20 years). Means and standard deviations were calculated separately for the two age strata overall and by sex.
To analyze age- and sex-specific distribution of each measurement, two-segment linear logistic regression models were estimated. These models assumed that the logarithm of the odds of having a measurement greater (versus less) than any constant c is linear from age 0 to 20 years and from age 20 onward. The probability of sex-specific intercepts and slopes was also considered. The models were fit using SAS PROC LOGISTIC, and model choice was based on Akaike’s information criterion.15 The advantage of these semiparametric models is that they generate age- and sex-specific prediction intervals directly, without having to make distribution assumptions. To corroborate these results qualitatively and to provide intuitively understandable growth rates, segmented linear regression was also performed, again with allowance for possible sex differences.
For subjects <20 years, the separate effects of age and body size on the LAA dimensions were assessed with multiple linear regression models, including various combinations of age, height, weight, and body surface area. The choice among models was guided by Mallows’ Cp statistics.16
Means and standard deviations of each measurement by sex and age (≥20 or <20 years) are shown in the Table⇓. Prediction intervals based on logistic models are shown in Figs 3⇓, 4⇓, and 5⇓. Age and sex accounted for only 27%, 28%, and 13% of the variation of LAA orifice size, length, and width in subjects younger than 20 years. In subjects ≥20 years, age and sex accounted for only 3%, 1%, and 2% of the variation of these three parameters.
Mean orifice size, length, and width increased at average rates of 0.024, 0.041, and 0.030 cm/yr, respectively, during the first 20 years of life (P<.01). Length and width increased at a slower rate in female subjects. In adults ≥20 years old, orifice size increased, length decreased, and width increased at average rates of 0.0016, −0.0040, and 0.0019 cm/yr, respectively. However, these average rates of change were due almost exclusively to the changes observed in male subjects, in whom the estimated rates were 0.0026, −0.0072, and 0.0037 cm/yr, respectively (based on selected models).
Effects of Body Size
When age, sex, and body size factors of height and weight or body surface area were considered simultaneously in the <20-year age group, the best model (based on Mallows’ Cp) for LAA diameter and length was based on age and sex. LAA width was associated slightly more strongly with height than with age, but this difference was not enough to warrant substituting height for age in the model. None of the body size measurements had an independent association with LAA dimensions. In the ≥20-year age group, body size did not make a significant contribution to prediction except that height tended to replace sex as a predictor of LAA width (P=.02). Overall, in both groups, body size made little additional contribution over age and sex in predicting LAA dimensions.
Fifty-four percent of LAAs had two lobes, and the number ranged between one and four lobes (Fig 6⇓). There were no age or sex differences.
Most LAAs (97%) had pectinate muscles ≥1 mm in width. Pectinate muscles <1 mm (3.0%) were noted only in the first and last decades. No sex-related differences were noted.
The LAA is a small, muscular extension of the left atrium arising near the left pulmonary veins. It arises anterolaterally and lies in the left atrioventricular sulcus atop the proximal portion of the left circumflex artery.
It has been described as a long, narrow, tubular, wavy, hooked appendage with a narrow junction and crenelated lumen.13 14 17 Differences between the LAA and the right atrial appendage have also been noted,13 14 17 18 and their importance in the assessment of situs and congenital heart malformations has been stressed. No previous study has defined the range of normal dimensions in a large number of normal hearts.
Unlike the right atrial appendage, in which the normal anatomy has been studied in 23 randomly selected patients,18 the LAA has been relatively ignored except for general observations. This study demonstrated age- and sex-related differences in LAA dimensions and firmly established that the LAA is multilobed (80% have two or more lobes). These lobes often lie in different planes.
Improved imaging techniques and the use of biplane19 and multiplane20 TEE have allowed visualization of the LAA, which previously was difficult to visualize by other imaging methods. Because accuracy of LAA thrombus detection with TEE is important in the precardioversion evaluation of patients with atrial fibrillation,9 10 21 it is vital to know what variations, especially in location and size of the pectinate muscle, exist in the normal anatomy of the LAA.
Mistaking a normal lobe for a thrombus or missing an LAA thrombus in a secondary lobe is possible if one is not aware of the variable anatomy of the LAA. This study demonstrated variability in the anatomy of the LAA and substantiated the existence of the multilobed LAA and thus the importance of searching for more than one lobe with TEE.
Larger pectinate muscles (≥1 mm) occur in 97% of LAAs and constitute another potential pitfall in TEE imaging of the LAA. Small (<1 mm) pectinate muscles were seen only in the first and last decades of life.
The accurate detection of thrombus and intra-atrial masses with TEE is important in the evaluation of patients with atrial fibrillation or stroke (or both). One must be aware of the complexity and variability of the shape, size, and number of lobes of the LAA and the size of the pectinate muscle to avoid misinterpretation. This study established the complexity of the LAA and age- and sex-related changes associated with it.
- Received April 9, 1997.
- Revision received May 27, 1997.
- Accepted June 6, 1997.
- Copyright © 1997 by American Heart Association
Pollick C, Taylor D. Assessment of left atrial appendage function by transesophageal echocardiography: implications for the development of thrombus. Circulation. 1991;84:223-231.
Pop G, Sutherland GR, Koudstaal PJ, Sit TW, de Jong G, Roelandt JR. Transesophageal echocardiography in the detection of intracardiac embolic sources in patients with transient ischemic attacks. Stroke. 1990;21:560-565.
Sharma S, Devine W, Anderson RH, Zuberbuhler JR. The determination of atrial arrangement by examination of appendage morphology in 1842 heart specimens. Br Heart J. 1988;60:227-231.
Gorman JW, Toman RJ. Selection of variables for fitting equations to data. Technometrics. 1966;8:27-51.
Tuccillo B, Stümper O, Hess J, van Suijlen RJ, Bos E, Roelandt JR, Sutherland GR. Transoesophageal echocardiographic evaluation of atrial morphology in children with congenital heart disease. Eur Heart J. 1992;13:223-231.