Does the left atrial appendage morphology correlates with the risk of stroke in patients with atrial fibrillation? Result from a multicenter study.

Size: px

Start display at page:

Download "Does the left atrial appendage morphology correlates with the risk of stroke in patients with atrial fibrillation? Result from a multicenter study."

Abel Stafford
6 years ago
Views:

1 Does the left atrial appendage morphology correlates with the risk of stroke in patients with atrial fibrillation? Result from a multicenter study. Luigi Di Biase, MD, PhD, Fiorenzo Gaita, MD, Ilaria Salvetti, MD, Sebastiano Gili, MD, Rodney Horton, MD, Matteo Anselmino, MD, PhD, Prasant Mohanty, MD, Pasquale Santangeli, MD, Javier E. Sanchez, MD, Andrea Natale, MD, FACC, FESC, FHRS Texas Cardiac Arrhythmia Institute at St. David s Medical Center, Austin, Texas, USA; Department of Biomedical Engineering, University of Texas, Austin, Texas, USA; Department of Cardiology, University of Foggia, Foggia, Italy; University of Turin, Turin, Italy dibbia@gmail.com

2 DISCLOSURES I am a consultant for Biosense Webster Hansen Medical

3 BACKGROUND The left atrial appendage (LAA) represents one of the major sources of cardiac thrombus formation responsible for TIA/stroke in patients with atrial fibrillation (AF).

4 Thrombus in Left Atrial Appendage Associated with Stroke Thrombus Thrombus in left atrial appendage is correlated with increased thromboembolic risk in AF Left Atrial Appendage Chimowitz. Stroke 1993; 24: 1015 Zabalgoitia. J Am Coll Cardiol 1998; 31: 1622

5 Etiology of Stroke in Atrial Fibrillation

6 AIM of the STUDY We quantitatively studied various morphologic parameters of the left atrial appendage (LAA) by computed tomography (CT), and magnetic resonance imaging (MRI) to categorize different LAA morphologies and we tried to correlate the morphology with the history of stroke/tia.

7 METHODS The study population consisted of 932 patients with drug refractory atrial fibrillation (AF) planning to undergo AF ablation. All patients underwent cardiac CT or MRI and care was taken to obtain LAA frames. All patients were screened for history of TIA/stroke. LAAs were categorized into different morphologies which included Chicken Wing, Windsock, Cauliflower and Cactus.

8 CT Cardiac MDCT and Segmentation Patients were scanned with contrast-enhanced ECG gated CT scan (Lightspeed Ultra, GE Healthcare, VA, USA). The slice acquisition thickness ranged from to 1.25 mm. Three-dimension structures of the left atrium (LA) and LAA, were constructed using the volume rendered postprocessing technique. The morphology of the LAA was also evaluated using multiplanar reconstruction.

9 CT The Classification of LAA Ostium Morphology The LAA ostium morphology was classified according to the shape of the orifice after the LAA was taken off LA in 3D images and the classification was verified by oblique transverse view of the LAA orifice. Based on it morphologies, the LAA was classified as: (1) The ChickenWing LAA is an anatomy whose main characteristic is an obvious bend in the proximal or middle part of the dominate lobe or folding back of the LAA anatomy on itself at some distance from the perceived LAA ostium. This LAA type may vary with or without secondary lobes or twigs, with the different measured distance to this bend as well as with the different orientation (anterior, superior, inferior, etc.) of the bend relative to the main lobe. (2) The WindSock LAA is an anatomy in which 1 dominant lobe of sufficient length is the primary structure. Variations of this LAA type arise with the location and number of secondary or even tertiary lobes arising from the dominate lobe in inferior direction.

10 CT The Classification of LAA Ostium Morphology (3) The Cauliflower LAA is an anatomy whose main characteristic is an LAA that has limited overall length with more complex internal characteristics. Variations of this LAA type are demonstrated by a more irregular shape of the perceived LAA ostium (oval vs. round), the number of significant lobes present and lack of 1 dominate lobe and the close proximity of internal separations or prominent pectinate ridges to the perceived LAA ostium. (4) The Cacuts LAA is an anatomy whose main characteristic is a dominant central lobe with secondary lobes extending from the central lobe in both superior and inferior directions. Variations of this type relate to the number, location and orientation of the secondary lobes. This type of LAA may present like a fork with a dominant lobe and with 2 or 3 secondary lobes at the top of LAA.

11 Image acquisition MRI All patients underwent a contrast-enhanced magnetic resonance imaging of the left atrium (LA) by intravenous administration of 0.2 mmol/kg of a paramagnetic contrast agent (Gadobutrolo, GADOVIST, Bayer S.P.A., Berlin, Germany) at a rate of 3.5 ml/s, followed by a bolus of 20 ml of physiological solution. Images were obtained with a body-array coil 1.5 Tesla magnetic resonance imaging system (Magneton Avanto 1.5T, Siemens, Erlangen, Germany). Three dimensional (3D) magnetic resonance angiography (MRA) was obtained with a breath-hold 3D fast-field Spoiled Gradient Echo (SPGR) imaging sequence performed in sagittal, coronal and axial views to obtain an anatomical view of the entire thorax. A narrow bandwidth of khz was used to reduce noise and improve the signal-to-noise ratio.

12 Image analysis MRI Global image thresholding was performed to obtain preliminary 3D, contrast visualization of the cavities (atrium, aortic arch, pulmonary veins and left ventricle). This preliminary reconstruction was then segmented in order to isolate structures of interest, in particular two reconstructions were obtained: the first including LA and left atrium appendage (LAA); the second separating LAA from LA at the auricolar ostium level, identified by endocardial and pericardial views, by a tangent cut. Referring to Veinot, et al., a lobe was considered a structure that: 1) is clearly identifiable from the outside, detaches itself from LAA principal body, and to which corresponds a completely distinct cavity; 2) is occasionally, but not necessarily, associated with the deviation in the orientation of the main LAA body.

13 MRI Measurements The following measurements were recorded at the ostium level of the LAA: perimeter, major and minor diameters (perpendicular lines), and ostial area. The LAA, based on a morphologic description, was included in on the four following types : Cactus: LAA body presents straight; the distance between the center of the LAA ostium and the most distal LAA body end was recorded. Chicken Wing: LAA body presents a single curve; three distances were recorded: between the center of the LAA ostium and the most distal LAA body (1); between the center of the LAA ostium and the first straight section (2); between end of distance 2 and the most distal LAA body end (3). Windsock: LAA body presents a double curve; four distances were measured: between the center of the LAA ostium and the most distal LAA body (1); between the center of the LAA ostium and the first straight section (2); between the first and the second straight sections (3); between the second straight section and the most distal LAA body end (4). Cauliflower: LAA body presents a T shape; three distances were recorded: between the center of the LAA ostium and the T-bifurcation long axis (1); the distance between the previous point and both ends (2 and 3) of the horizontal T segment.

14 CT scan types (1) The ChickenWing LAA is an anatomy whose main characteristic is an obvious bend in the proximal or middle part of the dominate lobe or folding back of the LAA anatomy on itself at some distance from the perceived LAA ostium. This LAA type may vary with or without secondary lobes or twigs, with the different measured distance to this bend as well as with the different orientation (anterior, superior, inferior, etc.) of the bend relative to the main lobe. 1 1

15 CT scan types (2) The WindSock LAA is an anatomy in which 1 dominant lobe of sufficient length is the primary structure. Variations of this LAA type arise with the location and number of secondary or even tertiary lobes arising from the dominate lobe in inferior direction. (3)The Cauliflower LAA is an anatomy whose main characteristic is an LAA that has limited overall length with more complex internal characteristics. Variations of this LAA type are demonstrated by a more irregular shape of the perceived LAA ostium (oval vs. round), the number of significant lobes present and lack of 1 dominate lobe and the close proximity of internal separations or prominent pectinate ridges to the perceived LAA ostium. (4)The Cacuts LAA is an anatomy whose main characteristic is a dominant central lobe with secondary lobes extending from the central lobe in both superior and inferior directions. Variations of this type relate to the number, location and orientation of the secondary lobes. This type of LAA may present like a fork with a dominant lobe and with 2 or 3 secondary lobes at the top of LAA

16 MRI types Cactus type Chicken wing type Cactus: LAA body presents straight; the distance between the center of the LAA ostium and the most distal LAA body end was recorded Chicken Wing: LAA body presents a single curve; three distances were recorded: between the center of the LAA ostium and the most distal LAA body (1); between the center of the LAA ostium and the first straight section (2); between end of distance 2 and the most distal LAA body end (3).

17 MRI types Windsock type Cauliflower type Windsock: LAA body presents a double curve; four distances were measured: between the center of the LAA ostium and the most distal LAA body (1); between the center of the LAA ostium and the first straight section (2); between the first and the second straight sections (3); between the second straight section and the most distal LAA body end (4). Cauliflower: LAA body presents a T shape; three distances were recorded: between the center of the LAA ostium and the T-bifurcation long axis (1); the distance between the previous point and both ends (2 and 3) of the horizontal T segment.

18 Chicken wing 3 2 1

19 Windsock

20 Cauliflower 2 3 1

21 Cactus

22 Statistical Analysis All continuous data are presented as mean ± standard deviation and were compared using analysis of variance (ANOVA) or Kruskal-Wallis test where appropriate. Categorical variables are described as count and percent and compared by using Pearson's chi-square or Fisher s exact test. Since classification of LAA categories was determined by operators using CT and MRI, it was necessary to test interoperator concordance. Cohen's Kappa was utilized to assess estimate inter-rater agreement.

23 Statistical Analysis Multivariable logistic model was used for identifying significant predictors of stroke/tia. All potential confounders were entered into the model based on known clinical relevance, or significant association observed in univariate analysis. The controlling variables used in the model were: age, gender, hypertension, diabetes, AF type and CHADS2 score. The odds ratio (OR) and 95% confidence interval (CI) of stroke/tia were computed. All tests were two-sided and a P-value <0.05 was considered statistically significant.

24 RESULTS: Baseline Characteristics Overall Population n=932 Type 1 (Cactus) n=278 Type 2 (Chicken Wing) n=451 Type 3 (Windsock) n=179 Type 4 (Cauliflower) n=24 P value Groups with pairwise sign difference Age, yrs 59±10 59±09 57±11 59±10 62± Male 734(79%) 218(78%) 356(79%) 147(82%) 13(55%) vs. 3; 4 vs. 2 AF Type PAF 548(59%) 167(60%) 266(59%) 100(56%) 15(64%) 0.81 PER 336(36%) 89(32%) 167(37%) 73(41%) 7(28%) LSP 48(5%) 22(8%) 18(4%) 6(3%) 2(8%) AF Duration, mo 59±65 67±77 30±41 50±62 47± BMI 27±04 27±04 27±04 27±03 26± Dyslipidemia 218(23%) 68(25%) 99(22%) 47(27%) 4(18%) Hypertension 450(48%) 143(52%) 201(45%) 95(53%) 11(45%) 0.15 CHF 42(5%) 9(3%) 19(4%) 13(7%) 1(4%) Diabetes 40(4%) 19(7%) 13(3%) 6(4%) 2(9%) Prior stroke/tia 78(8%) 35(12%) 20(4%) 19(10%) 4(18%) < vs. 1; 2 vs. 3 CAD 45(5%) 15(5%) 24(5%) 4(2%) 2(9%) CHADS (46%) 115(42%) 237(53%) 67(37%) 9(36%) < vs. 3; 2 vs. 1 CHADS (40%) 111(40%) 173(38%) 84(47%) 9(36%) CHADS (14%) 51(19%) 41(9%) 28(16%) 7(27%) < vs. 1; 2 vs. 4 LV EF, % 60±07 60±08 59±07 60±07 60± ARB 143(15%) 46(17%) 52(11%) 41(23%) 4(18%) ACE Inhibitor 150(16%) 47(17%) 67(15%) 32(18%) 4(18%) 0.16 Beta-blocker 304(33%) 90(32%) 147(33%) 56(31%) 11(45%) 0.56 Aspirin/Plavix 153(16%) 45(16%) 76(17%) 30(17%) 2(9%) Lipid-lowering 60(6%) 28(10%) 19(4%) 13(7%) 0(0%) vs. 1 LAA volume 14.3± ±07.58 // 14.9±06.71 // LAA Velocity, mm 74.5± ± ± ± ± LAA AP Diam, mm 45.4± ± ± ± ± LAA Long Diam 60.7± ± ± ± ± LAA Lat-Median Diameter 46.2± ± ± ± ±

25 RESULTS No statistically significant bias was noted in classifying LAA morphology by operators using CT and MRI (Kappa=0.667; 95% CI , p= 0.001).

26 RESULTS: Baseline by Prior Stroke Status No Hx of Stroke/TIA Prior Stroke/TIA (n-854) (n=78) P value Age, yrs 58±10 62± Male 674(79%) 60(76%) LAA Type Cactus 243(28%) 35(44%) Chicken Wing 431(50%) 20(26%) <0.001 Windsock 160(19%) 19(24%) Cauliflower 20(2%) 4(5%) BMI 27±04 27± Dyslipidemia 193(23%) 25(32%) Hypertension 409(48%) 41(53%) CHF 39(5%) 3(4%) Diabetes 35(4%) 5(6%) CAD 43(5%) 2(3%) 0.33 CHADS2= 0 426(50%) 2(3%) <0.001 CHADS2= 1 372(44%) 5(6%) <0.001 CHADS2 2 56(7%) 71(91%) <0.001 LV EF, % 58±08 60± ARB 129(15%) 14(18%) ACE Inhibitor 129(15%) 21(26%) Beta-blocker 281(33%) 23(29%) Aspirin/Plavix 148(17%) 5(6%) Lipid-lowering therapy 51(6%) 9(12%) LAA volume 14.13± ± LAA Velocity, mm 74.77± ± LAA AP Diameter, mm 45.36± ± LAA Long Diam, mm 60.64± ± LAA Lat-Median Diam, mm 46.27± ±

27 Univariate Analysis: Odds ratio Error bars display 95% C.I. of odds ratio

28 Multivariable Analysis After controlling for CHADS2 score, gender, and AF type Chicken Wing was found to be more likely to remain stroke-free (odds ratio 19, p= 0.043). Compared to chicken wing: Cactus had 4 times (OR 4.1, 95%, p= 0.046), Windsock- 5 times (OR 4.8, p=0.038), and Cauliflower 8 times (OR 8.0, p=0.056) more likely to have prior stroke/tia

CONCLUSIONS This study suggests that patients with Chicken Wing morphology are less likely to have an embolic event even after controlling for

29 CONCLUSIONS This study suggests that patients with Chicken Wing morphology are less likely to have an embolic event even after controlling for comorbidities. If confirmed, these results could have a relevant impact on the anticoagulation management of patients with an intermediate risk for stroke

30 Does the left atrial appendage morphology correlates with the risk of stroke in patients with atrial fibrillation? Result from a multicenter study. Luigi Di Biase, MD, PhD, Fiorenzo Gaita, MD, Ilaria Salvetti, MD, Sebastiano Gili, MD, Rodney Horton, MD, Matteo Anselmino, MD, PhD, Prasant Mohanty, MD, Pasquale Santangeli, MD, Javier E. Sanchez, MD, Andrea Natale, MD, FACC, FESC, FHRS Texas Cardiac Arrhythmia Institute at St. David s Medical Center, Austin, Texas, USA; Department of Biomedical Engineering, University of Texas, Austin, Texas, USA; Department of Cardiology, University of Foggia, Foggia, Italy; University of Turin, Turin, Italy dibbia@gmail.com

Role of cardiac imaging for catheterbased left atrial appendage closure

Role of cardiac imaging for catheterbased left atrial appendage closure Ana G. Almeida, MD, PhD Cardiology University Hospital Santa Maria, Lisbon Ana G. Almeida, MD, PhD, FESC University Hospital Santa