Page 20 - Journal of Structural Heart Disease Volume 3, Issue 1

13
Original Scienti c Article
ed LAAs, with one half showing a moderate decrease and the other showing a severe decrease in angle (-30 to -90°). This combination of unusual direction and change in curvature may complicate deep insertion of the WATCHMAN device into the LAA. Our charac- terization of LAA morphology could improve delivery system design to facilitate access to the LAA and may provide guidance for catheter tip con guration, al- lowing smoother positioning within the LAA. These improvements have the potential to reduce the risk of perforation and pericardial hemorrhage. Optimal- ly, multiple delivery sheaths with a variety of curva- tures that approximate angulations of the append- age assessed with non-invasive imaging such as CT or MRI together with pre-procedural characterization of the LAA with respect to its angulation and curvature would minimize the need for undue sheath manipu- lation and device recapture maneuvers.
LAA Lobes and Lobules
Complete LAA assessment and closure is compli- cated by multi-lobed appendages, as lobes can exist in di erent planes requiring complex visualization in multiple views. Our results regarding the number of lobes are consistent with those of Heist et al. [10], who also used MRI to analyze the LAA in AF patients and found that about two-thirds of LAAs had one lobe, one-third had two lobes, and a small proportion (<5%) had multiple lobes (≥3 lobes). However, Veinot et al. [9] found a much higher prevalence of multiple lobes in post-mortem specimens of normal hearts, with 80% of LAAs having more than one lobe. These discrepancies among studies may arise from di er- ences in lobe de nitions and a lack of distinction be- tween lobes and smaller variants (i.e., lobules). In ad- dition, further extra- and intraluminal criteria (i.e., an external crease and accessibility by a probe, respec- tively) potentially increasing the number of lobes were taken into account by Veinot et al. but could not be assessed in our study due to di erent analytical methods. Finally, identi cation of small lobules may have been more di cult in our study due to smaller spatial resolution by non-gated MRI compared with direct visualization in post-mortem studies.
Limitations of our study include a small sample size; therefore, quoted proportions should be regard-
ed as approximate. However, our study was undertak- en to demonstrate the range of shapes and angula- tions that can occur in LAAs to avoid the assumption of uniform geometry rather than to report precise proportions of each morphology. Twenty-seven pa- tients with inadequate image quality for LAA assess- ment were excluded, which in theory may have led to selection bias. However, such a bias would have required that LAA morphology a ected CMR image quality, which is unlikely as this is not dependent on LAA orientation. Most importantly, we performed our analysis based on the assumption that knowledge of LAA morphology and recognition of its variants may lead to improvements in device and delivery sys- tem designs, thereby reducing the risk of procedur- al complications. Though plausible to operators, this assumption remains to be proven. Our study popula- tion consisted only of patients with current/paroxys- mal AF, who may likely have larger left atria. Although this could alter LAA shape, signi cant modi cations seem unlikely. Furthermore, these patients are also those in whom TOE assessment for thrombus or LAA device closure is usually performed; therefore, they are a valid group to study.
Our nding that the morphology and orienta- tion of the LAA can vary signi cantly may have im- plications for clinical practice. Retroverted LAAs and multiple LAA lobes are common, and the shape and curvature/angulation displays wide variation. Knowl- edge of these morphological aspects may improve the detection of LAA thrombus on TOE and may also improve delivery sheath con guration and LAA oc- clusion device design to reduce procedural risk.
Acknowledgements
Saul G Myerson is funded by the Oxford National Institute for Health Research (NIHR) Biomedical Re- search Centre Program.
Con ict of Interest
The authors have no con ict of interest relevant to this publication.
Comment on this Article or Ask a Question
Joy, S. et al.
LAA Morphology in Non-Valvular AF