Int J Cardiovasc Imaging (2015) 31:1191–1199 DOI 10.1007/s10554-015-0675-1
ORIGINAL PAPER
Relation between left atrial wall composition by late gadolinium enhancement and complex fractionated atrial electrograms in patients with persistent atrial fibrillation: influence of non-fibrotic substrate in the left atrium Sung Ho Hwang1 • Yu-Whan Oh1 • Dae In Lee2 • Jaemin Shim2 • Sang-Weon Park2 Young-Hoon Kim2,3
•
Received: 14 March 2015 / Accepted: 6 May 2015 / Published online: 10 May 2015 Ó Springer Science+Business Media Dordrecht 2015
Abstract The complex fractioned atrial electrogram (CFAE) has been considered as the catheter ablation target of left atrium (LA) under persistent atrial fibrillation (PeAF). We evaluated the relation between the LA wall composition by late gadolinium enhancement cardiac magnetic resonance (LGE-CMR) and the CFAE in patients with PeAF. Forty-three patients underwent LGE-CMR and CFAE mapping before catheter ablation for PeAF. The LA wall substrates were classified into three: the fibrotic, intermediate, and normal substrates by using two thresholds of 2 standard deviation (2-SD) and 6-SD above the mean signal from the normal myocardium. For each of 12 preselected LA wall regions, the composition ratios (CRs) of fibrotic, indeterminate, and normal substrates were calculated as a percentage to the volume of LA wall region, and compared depending on the CFAE, respectively. The CR of normal substrate was significantly greater at the LA wall region with CFAE (52 ± 38 % vs. 20 ± 28 %, P \ 0.01) than without CFAE. In contrast, the LA wall region with CFAE showed significantly lower CRs of intermediate substrate (39 ± 34 % vs. 57 ± 31 %, P \ 0.01) and fibrotic substrate (7 ± 17 % vs. 21 ± 24 %, P \ 0.01) than did the LA wall region without CFAE, respectively. Thus, the high CR ([18 %) of normal
& Young-Hoon Kim [email protected] 1
Department of Radiology, Korea University Medical Center, Seoul, Republic of Korea
2
Division of Cardiology, Department of Internal Medicine, Korea University Medical Center, Seoul, Republic of Korea
3
Division of Cardiology, Department of Internal Medicine, Anam Hospital, Korea University, Anam-dong 5-ga, Seongbuk-gu, Seoul 136-705, South Korea
substrate predicted the CFAE at the corresponding LA wall region with 71 % sensitivity and 62 % specificity. In conclusion, the evaluation of LA wall normal substrate by LGE-CMR might be useful to predict the CFAE occurrence before catheter ablation of PeAF. Keywords Left atrium Atrial fibrillation Fibrosis Myocardium Magnetic resonance imaging
Introduction Catheter ablation for electrical isolation of arrhythmogenic substrate has gained wide acceptance for the management of refractory atrial fibrillation (AF) [1, 2]. The pulmonary veins (PVs) have been considered as the main drivers of persistent AF (PeAF), and the electrical isolation of PVs can help achieve restoration of sinus rhythm in patients with PeAF [3]. Beyond the PVs, the complex fractionated atrial electrogram (CFAE) has been proposed as an important left atrial (LA) phenomenon maintaining PeAF. Catheter ablation strategies targeting the LA wall associated with the CFAE in addition to isolation of PV have been applied to improve the maintenance of sinus rhythm in patients with refractory PeAF [4, 5]. The CFAE is defined as complex electrograms with two or more deflections, continuous activity, and cycle length (CL) B120 ms [5]. Delineation of transient CFAE needs an invasive catheter-based electrophysiologic study. To minimize the invasive procedures, prior studies have evaluated the CFAE by using noninvasive methods such as computed tomography and magnetic resonance systems [6, 7]. Several studies about the LA wall substrates have suggested that PeAF is associated with progressive electrical and structural LA wall remodeling [8–10]. It has also
123
1192
been known that the summation of electrograms from overlying layers of normal myocardial fibers, autonomic nerves, and rotors might be related to the CFAE [11, 12]. Thus, we believe that a detailed assessment of LA wall composition may improve the prediction of transient CFAE occurrence. Late gadolinium enhancement cardiac magnetic resonance (LGE-CMR) has been considered a gold standard modality for differentiation between the fibrosis and normal myocardium in various cardiovascular diseases [13, 14]. Jadidi et al. [6] reported that the patchy LA fibrosis detected with LGE-CMR is inversely associated with the CFAE occurrence. So, we could surmise that the non-fibrotic substrates might also be associated with the CFAE occurrence in the LA wall. In the present study, we evaluated the LA wall composition of various substrates determined by LGE-CMR, depending on the CFAE map results. Furthermore, we tried to determine the feasibility of LGE-CMR in the prediction of CFAE occurrence in patients with PeAF.
Methods Study population Between July 2013 and July 2014, 65 patients sequentially underwent preprocedural cardiac magnetic resonance (CMR) examination for the evaluation of the LA and PVs, and CFAE mapping before catheter ablation for PeAF at the Korea University Anam Hospital. The patient with PeAF had an atrial fibrillation episode that lasted longer than 7 days or required termination by cardioversion without catheter ablation. In this study, the exclusion criteria were contraindication to CMR examination (e.g., severe claustrophobia, automatic implantable defibrillators, or pacemakers) and unacceptable quality of CMR images Eventually, of all selected 65 patients, 22 (33.8 %) were excluded due to the unacceptable quality of LGE-CMR images in the quantification of LA wall composition. A total of 43 patients (35 men; mean age, 54.4 ± 9.6 years) were enrolled into the present study. During CMR data acquisition, the mean heart rate of all enrolled patients was 69.2 ± 15.2 beats/min. No patients showed AF attack during the CMR examination. This retrospective study was approved by the institutional review board. All patients provided written informed consent for their participation. CMR protocol Before catheter ablation, CMR examination was performed by using a 3-T magnetic resonance system (Achieva; Philips Medical Systems, Best, the Netherlands) with a
123
Int J Cardiovasc Imaging (2015) 31:1191–1199
32-element phased-array cardiac coil. First, a three-dimensional (3D) spoiled gradient-echo contrast-enhanced timing robust angiography (CENTRA) was performed to define the cardiac anatomy of the LA and the PVs after the injection of 0.2 mmol/kg gadolinium contrast (Dotarem; Guerbet, S.A., Villepinte, France). Then, high spatial resolution LGE-CMR images were acquired approximately 15–20 min after the administration of the gadolinium contrast by using a 3D inversion recovery-prepared, respiration-navigated, electrocardiography (ECG) gated, gradient-echo pulse sequence after the Look–Locker sequence to identify the optimal nulling time for the normal left ventricular (LV) myocardium. Typical data acquisition parameters for LGE-CMR were free breathing by using a respiratory navigator with a 6-mm acceptance window; a transverse imaging volume with a voxel size of 1.5 9 1.5 9 1.5 mm; TR/TE of 4.7/1.4 ms; TI of 230–270 ms; flip angle of 25°; bandwidth of 127 Hz/pixel; 1 R wave to R wave (RR) interval between inversion pulses; phase encoding in the right–left direction; and parallel imaging with the sensitivity encoding (SENSE) technique with R = 2. The LGE-CMR images were acquired within a window of 100–150 ms for each RR wave interval. Respiratory navigator inflow artifacts were reduced by lowering the navigator rescales and positioning the navigator away from the right-sided PV. Depending on the successful leading navigator placement, free-breathing images aimed for a navigator efficiency of [30 %. Then, all LGE-CMR images were transferred into the workstation for image analysis. Analysis of LGE-CMR images All LGE-CMR images were analyzed independently by two experienced radiologists (with 7 and 20 years of experience in cardiac image interpretation) blinded to the patients’ clinical and electrophysiological data, by using a commercial software workstation (Terarecon iNtuition; TeraRecon, Foster City, CA, USA). Figure 1 shows the 3D reconstruction process of LGECMR images. On transverse or coronal LGE-CMR images, the epicardial and endocardial borders of the LA wall were semiautomatically contoured to select the entire LA wall and reconstruct a 3D LA model. In the quantitative analysis of the selected LA wall, we used signal threshold methods involving the manual delineation of regions of interest (ROI) that included, to the greatest possible extent, the normal left ventricular (LV) wall of nulled signal intensity on LGE-CMR image. On the basis of the mean and standard deviation (SD) values from ROI, the 6-SD and 2-SD thresholds above the mean signal in the normal LV wall were calculated [15, 16] (Fig. 1). By using the 6-SD and 2-SD thresholds, all LA wall substrates were classified into
Int J Cardiovasc Imaging (2015) 31:1191–1199
1193
Fig. 1 Determination of the normal, intermediate, and fibrotic substrates in the left atrial (LA) wall by using LGE-CMR. The 2-standard deviation (SD) and 6-SD threshold algorithms are established from the left ventricular myocardium of nulled signal intensity on the LGE-CMR image (a). The epicardial and endocardial borders of LA wall (green color) on LGE-CMR images (b, c) are semiautomatically selected for the reconstruction of the 3D LA wall
model (d). In the characterization of each of different LA wall substrates, the fibrotic substrates with signal intensities (green color) equal to or greater than the 6-SD threshold are selected in the posterior LA wall (white arrow) on transverse LGE-CMR image (e). In addition, the normal substrates with signal intensities (green color) less than the 2-SD threshold are selected in the LA roof (black arrow) on coronal LGE-CMR image (f)
three: (1) the fibrotic substrate with signal with [6-SD threshold; (2) the intermediate substrate with a signal of C2-SD threshold and \6-SD threshold; and (3) the normal substrate with signal of \2-SD threshold. Furthermore, different color look-up table masks were applied to the yellow for the fibrotic substrate, the gray for the indeterminate substrate, and the blue for the normal substrate on the 3D LA model for better delineation of LA wall composition (Figs. 1, 2).
system; St. Jude Medical, St. Paul, MN, USA) was generated. Bipolar intracardiac electrograms were filtered between 30 and 300 Hz. Intracardiac recordings were simultaneously recorded with CARTO and a computerized multichannel recording system (EP Med Systems Inc., Mt. Arlington, NJ, USA). Tachycardia and CL were monitored and recorded from the mapping catheter for the LA wall. The CFAE was defined as atrial electrograms with CL B120 ms [5]. For a CFAE map, as previously described, contact bipolar electrograms were obtained during an atrial fibrillation episode lasting[3 min; they were measured for more than 6 s at each LA wall site and were approximately equally distributed within the LA wall. The mean CL of the local electrogram was projected onto the LA wall shell with a color-coded display (Fig. 3). Additionally, the LA wall pixels of the CFAE were coded as red or white in the CFAE map.
CFAE mapping For recording and stimulation, a duodecapolar catheter (St. Jude Medical, St. Paul, MN, USA) and a decapolar catheter (Bard Electrophysiology, Lowell, MA, USA) were placed in the coronary sinus and in the low right atrium (RA) and high RA through the left femoral vein, respectively. A quadripolar catheter was also placed in the superior vena cava. After double transseptal puncture, the patients were administered with anticoagulants such as intravenous heparin to maintain an activated clotting time of between 300 and 400 s. Then, a 10-bipole or 20-bipole double-spiral circular mapping catheter (Lasso; Biosense Webster, Diamond Bar, CA, USA) was used for electrophysiologic mapping. Then, the CMR angiography merged electrophysiologic map (Ensite NavX
Comparison between LGE-CMR and CFAE map The entire LA wall was subdivided into 12 preselected LA wall regions: 6 on the posterior wall, and 6 on the anterior wall and septum (Fig. 4). For each of 12 LA wall regions, a quantitative analysis was performed to correlate the LGECMR images and CFAE map (Fig. 5). To determine the composition ratio (CR) of LA wall for each LA wall
123
1194
Int J Cardiovasc Imaging (2015) 31:1191–1199
Fig. 2 Three-dimensional (3D) reformatting and color coding of LA wall in basis of LGE-CMR for the characterization of left atrial (LA) wall composition. The normal substrate (blue), intermediate substrate (gray), and fibrotic substrate (red) can be simultaneously separated from the 3D LA wall model, transverse and coronal LGECMR images. Transverse LGE image shows a predominance of normal substrate in the anterior LA wall at the aortic valve level (arrow). Coronal LGE image shows a predominance of normal substrate in the LA roof (arrow)
Fig. 3 Anteroposterior view and left oblique view of the CFAE map. The LA wall pixels showing a mean cycle length (CL) of B120 ms are coded as red or white, presenting the CFAE occurrence
region, the summation of entire LA wall substrates was considered as the total volume of LA wall region. The mean total volume of every LA wall region was
123
3.1 ± 0.9 cm3. Furthermore, each CR of the normal substrate, the intermediate substrate, and the fibrotic substrate was also calculated as a percentage of the total volume of
Int J Cardiovasc Imaging (2015) 31:1191–1199
1195
Fig. 4 Schematic representation of the 12 preselected LA regions on the LGE-CMR image. On the posterior and anterior views of the 3D LA wall model, the 12 preselected LA wall regions include 6 regions at posterior LA wall (blue), 2 regions at LA septum (green), and 4 regions at anterior LA wall (yellow)
Fig. 5 Correlation between the LGE-CMR image and the CFAE map. The 3D LA wall model by LGE-CMR shows the focused distribution of the normal substrate (blue) in the anterior LA wall (white arrow) and posterior LA wall (black arrow), that well corresponds to the CFAE (red) with a cycle length of B120 ms on the CFAE map. In contrast, the LA septum with dense and large fibrotic substrates (yellow) on 3D LA wall model shows no CFAE
the LA wall at the corresponding LA wall region, respectively. In the analysis of LGE-CMR by the 2 investigators, intraobserver and intrerobserver reproducibility for the volume measurement of each LA wall substrates were calculated and found to be reliable (intraclass correlation coefficient [0.7 for all volume measurements). In the
evaluation of the CFAE within the corresponding LA region, the visual extent of the CFAE was evaluated by two blinded reviewers (two experts in electrophysiologic study) with consensus. Depending on the visual presence of the red or white-coded CFAE, every LA wall region was classified into two: (1) the CFAE region (with presence of
123
1196
Int J Cardiovasc Imaging (2015) 31:1191–1199
the white or red pixels in the corresponding LA wall region) and (2) the non-CFAE region (without the white or red pixels in the corresponding LA wall region). Statistical analysis Statistical analysis was performed by using SPSS software (version 19.0; IBM, Somers, NY, USA). All continuous data were expressed as mean ± SD, and all categorical data were presented as absolute values and percentages. Continuous variables were compared by using Student’s ttest depending on the CFAE map results such as the CFAE and non-CFAE at the preselected 12 LA wall regions. Receiver operating characteristic (ROC) curves were used to determine the accuracy of a variable in predicting the CFAE. A P value of \0.05 was considered statistically significant.
Results Table 1 shows the baseline clinical characteristics of the 43 enrolled patients. All patients had no diagnosis of coronary artery disease and cardiomyopathy. The mean LV ejection fraction was 54.2 ± 7.1 %. When considering the clinical history (e.g., hypertension, diabetes, heart failure, and stroke), the mean CHADS2 score [17] was 0.4 ± 0.8. Table 2 summarizes the frequency of CFAE and the CRs of three different LA wall substrates by LGE-CMR at every LA wall region. A high frequency ([50 %) of CFAE was noted in the posterior region (n = 22, 51 %), the superior septum region (n = 26, 60 %), the anterior roof region (n = 26, 60 %), and the appendage region (n = 26, 60 %). Additionally, the superior septum region, anterior roof region, and appendage region were located in the septum and anterior wall of the LA, which showed high mean CR of normal substrates [50 %. In contrast, the mean CR of fibrotic substrate was \5 % at the LA wall regions related to a high frequency ([50 %) of CFAE.
Table 1 Patients’ demographics (N = 43) Male
35 (81.3)
Age (years)
54.4 ± 9.6
Old age ([75 years)
0 (0)
Congestive heart failure
4 (9.2)
Hypertension
11 (25.5)
Diabetes mellitus
2 (4.6)
Stroke
2 (4.6)
CHADS2 score
0.4 ± 0.8
Data are presented as number (percentage) or mean ± standard deviation
123
When a total of 516 LA wall regions were evaluated, 152 (29 %) were assigned into the CFAE region. The mean CR of the normal substrate at the CFAE region was significantly greater than that at the non-CFAE region (52 ± 38 % vs. 20 ± 28 %, P \ 0.01) (Fig. 6). In contrast, the mean CR of the intermediate substrate at the CFAE region was significantly lower than at the non-CFAE region (39 ± 34 % vs. 57 ± 31 %, P \ 0.01) (Fig. 6). In addition, the mean CR of the fibrotic substrate at the CFAE region was also significantly lower than that at the nonCFAE region (7 ± 17 % vs. 21 ± 24 %, P \ 0.01) (Fig. 6). To establish the feasibility of the prediction of the CFAE region depending on the CR of each LA wall substrate by LGE-CMR, the sensitivity and specificity of various cutoffs related to the CR of each LA wall substrate were evaluated by using the ROC curve (Fig. 6). The areas under the ROC curves (AUCs) were 0.71 (P \ 0.01) for the CR of the normal substrate, 0.34 (P \ 0.01) for the CR of the intermediate substrate, and 0.31 (P \ 0.01) for the CR of the fibrotic substrate. When the CR of normal substrate was [18 % of the total volume of the LA wall at the preselected LA wall region, we could predict the CFAE occurrence at the corresponding LA wall region with a sensitivity of 71 % and a specificity of 62 % (Fig. 7).
Discussion We evaluated the LA wall composition by using LGECMR depending on the CFAE occurrence at each of the 12 preselected LA wall regions covering the entire LA wall in patients with PeAF. Among the 12 preselected LA regions, the posterior, superior septum, appendage, and anterior roof regions showed a high probability of CFAE occurrence. In the comparison of LA wall composition between the CFAE region and the non-CFAE region, the CFAE region showed a greater CR of the normal substrate, but lower CR of the fibrotic and intermediate substrates than did the non-CFAE region. Furthermore, we found that the high CR ([18 % of total LA wall volume at the preselected LA wall region) of the normal zone might be associated with the potential of CFAE at the corresponding LA wall region in patients with PeAF. LGE-CMR has been emerging as a powerful technique for identifying the hyperenhancement of fibrosis even in the thin LA wall [18–20]. Generally, patients with more remodeled LA wall have more extensive LA fibrosis than those with less remodeled LA wall [20, 21]. An appreciable correlation between LA fibrosis and low voltage (\0.5 mV) has been demonstrated [21, 22]. MalcolmeLawes et al. [23] revealed that the degree of hyperenhancement on LGE-CMR image correlates with the low tissue voltage representing the degree of fibrosis or injury.
Int J Cardiovasc Imaging (2015) 31:1191–1199 Table 2 Mean composition ratio (%) of LA wall substrate and the frequency of CFAE for each preselected LA wall region in 43 patients with persistent atrial fibrillation
1197
Normal
Intermediate
Fibrotic
CFAE occurrence
Posterior wall of the left atrium 1. Left superior PV region
12 ± 10
53 ± 21
36 ± 18
2 (5)
2. Posterior roof region
16 ± 21
78 ± 22
6 ± 11
13 (30)
3±5
76 ± 26
21 ± 26
2 (5)
4. Left inferior PV region
3. Right superior PV region
13 ± 9
70 ± 20
17 ± 18
2 (5)
5. Posterior region
49 ± 25
48 ± 24
3±6
22 (51)
6. Right inferior PV region
10 ± 8
69 ± 24
21 ± 21
2 (5)
Septum and anterior wall of the left atrium 7. Superior septum region
66 ± 32
34 ± 31
1±2
26 (60)
8. Anterior roof region
85 ± 9
12 ± 9
2±4
26 (60) 26 (60)
9. Appendage region
51 ± 44
48 ± 43
1±3
10. Inferior septum region
11 ± 13
29 ± 14
60 ± 6
4 (9)
11. Anterior region
11 ± 24
72 ± 33
17 ± 25
11 (26)
12. Mitral valve region
36 ± 40
40 ± 34
24 ± 27
17 (40)
Numbers in parentheses are percentages PV pulmonary vein
Fig. 6 Comparison of LA wall composition depending on the CFAE map result. There were significant differences in the composition ratio of each LA wall substrate (e.g., fibrotic substrate, intermediate substrate, and normal substrate) between the CFAE and the non-CFAE
By using LGE-CMR for tissue characterization, the ‘‘normal substrate’’ of baseline low signal intensity should first be defined because the fibrotic substrate and the intermediate substrate of hyperenhancement on LGE-CMR are identified on the basis of the relative signal difference from the ‘‘normal substrate’’ [13]. Signal intensity at 2-SD above the mean reference myocardium is the minimum threshold to differentiate between LGE and normal myocardium [16]. Despite the improved resolution of LGE-CMR image, the thin LA wall itself may sometimes disrupt the accurate segmentation of ‘‘normal substrate’’ on LGE-CMR images. Therefore, prior studies have used the high signal threshold based on references such as LA blood and LV wall for tissue characterization [23, 24]. Furthermore, the 6-SD
signal threshold of LGE can be feasible to define the LA wall corresponding to the low-voltage tissue on the electroanatomic map [24]. In the present study, the extreme signal thresholds by using the 2- SD and 6-SD thresholds based on the normal LV myocardium of nulled signal intensity were applied to define both the normal substrate and the fibrotic substrate. We believe that the classification of LA wall composition (e.g., normal, intermediate, and fibrotic substrates) according to the 2-SD and 6-SD threshold methods is somewhat applicable for the reproducible evaluation of the thin LA wall. The CFAE are not symmetrically located but can be predictably found in the certain LA wall regions, including the interatrial septum, posterior wall, anterior wall, roof,
123
1198
Fig. 7 Prediction of the CFAE in terms of the fibrotic, intermediate, and normal substrates by using the receiver operating characteristic (ROC) curve. At the preselected LA wall region, the volume of the normal substrate determined with LGE-CMR showed an area under the ROC curve (AUC) of 0.71. In addition, the high volume ([18 %) of normal substrate at the LA wall region predicted the CFAE occurrence with a sensitivity of 71 % and a specificity of 62 %
appendage base, and coronary sinus [5, 25]. In our study results, we also found that the LA wall regions frequently associated with the CFAE include the posterior, superior septum, appendage, and anterior roof regions of LA wall. In basis of our study results, the high frequency of CFAE in the septum and anterior wall of the LA was considerably compatible with the results from previous studies [5, 25]. In the aspect of anatomy, the anterior roof region and the appendage region include the overlapping layers of myocardial fibers such Bachmann’s bundle [11]. Additionally, the superior septum region where myocardial fibers from the left atrium and right atrium have specific muscular bundles has a high potential of CFAE [26]. Eventually, the ‘‘interatrial conduction’’ of muscular bundles or myocardial fiber may create CFAE with conduction abnormalities in the anterior roof, appendage, and superior septum regions. Thus, in the present study, we suggest that the ‘‘normal’’ substrate by using LGE-CMR might represent the presence of myocardial fiber, probably associated with the CFAE in the LA wall. Furthermore, the identification of LA wall composition by LGE before catheter ablation may provide an opportunity to characterize the PeAF pathophysiology. On the basis of our study results, the additional CFAE study should be recommended in PeAF patients with a high composition ratio of normal substrate by LGE especially in the anterior roof, appendage, and superior
123
Int J Cardiovasc Imaging (2015) 31:1191–1199
septum regions of LA wall. Substrate modification strategies for the CFAE have become increasingly used as part of ablation procedures in patients with PeAF [2, 5]. Although the mechanism of the CFAE is still unclear, the CFAE has been considered as the dyssynchronous activation of separate cell groups at pivot points, or the repetitive activations of the atrial fibrillation driver or local reentry circuit [27]. Nademanee et al. [5] reported that patients with PeAF have greater numbers and locations of CFAE in the LA wall than those with paroxysmal AF. In contrast, a prior study revealed that the LA wall area displaying the CFAE was smaller in long-lasting AF patients with a more electroanatomically remodeled LA wall than in those with less remodeled LA wall [28]. Jadidi et al. [6] demonstrated that the large fibrotic substrate detected with LGE-CMR is inversely associated with the CFAE occurrence in the LA wall. When we evaluated the various LA wall substrates of the low signal intensity or hyperenhancement by LGECMR, we also found that the fibrotic substrate of definite LGE was not a predominant component at the CFAE region of LA wall. Additional research is necessary to determine whether the prediction of CFAE occurrence by LGE-CMR may help the achievement of long-term success in the catheter ablation of PeAF. This study had several limitations. First, the sample size was small, and did not include a normal control group without a history of AF. A large follow-up study will be needed to define the clinical significance of noninvasive prediction of CFAE by using LGE-CMR before catheter ablation for PeAF. Second, the LGE-CMR resolution is limited for the evaluation of the thin LA wall. In this study, thin LA wall with a wall thickness of 1 mm may not be delineated or may be represented as the intermediate substrate with the resolution of the currently used LGE-CMR technique. Nevertheless, the reproducibility of measurement of the LA wall composition by using LGE-CMR was statistically reliable. With the advances in the LGE-CMR technique, it would be interesting to quantify the various LA wall substrates for evaluation of LA wall composition. Last, detailed studies with CFAE maps will be necessary to compare the CFAE map result and the LA wall signal intensity directly in the corresponding pixel points, because the visual assessment is limited for the accurate registration between the LGE-CMR and CFAE map. In summary, this study result emphasizes that the assessment of LA wall composition by using LGE-CMR may be useful to understand the mechanisms of CFAE in patients with PeAF. The fibrotic or intermediate substrates showing gray or bright signal intensity on LGE-CMR image might not sufficiently indicate a risk potential of CFAE in the preselected LA wall region. In contrast, the normal substrate of nulled signal intensity on LGE-CMR image might be associated with the CFAE occurrence in the
Int J Cardiovasc Imaging (2015) 31:1191–1199
preselected LA wall region under PeAF. Thus, the quantification of LA wall composition by using LGE-CMR can be useful for the prediction of the CFAE pattern in patients who will undergo catheter ablation for PeAF. Conflict of interest
None.
References 1. Haissaguerre M, Sanders P, Hocini M et al (2005) Catheter ablation of long-lasting persistent atrial fibrillation: critical structures for termination. J Cardiovasc Electrophysiol 16(11):1125–1137 2. Oral H, Chugh A, Yoshida K et al (2009) A randomized assessment of the incremental role of ablation of complex fractionated atrial electrograms after antral pulmonary vein isolation for long-lasting persistent atrial fibrillation. J Am Coll Cardiol 53(9):782–789 3. Haissaguerre M, Jais P, Shah DC et al (1998) Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med 339(10):659–666 4. Jais P, Haissaguerre M, Shah DC et al (1996) Regional disparities of endocardial atrial activation in paroxysmal atrial fibrillation. Pacing Clin Electrophysiol 19(11 Pt 2):1998–2003 5. Nademanee K, McKenzie J, Kosar E et al (2004) A new approach for catheter ablation of atrial fibrillation: mapping of the electrophysiologic substrate. J Am Coll Cardiol 43(11):2044–2053 6. Jadidi AS, Cochet H, Shah AJ et al (2013) Inverse relationship between fractionated electrograms and atrial fibrosis in persistent atrial fibrillation: combined magnetic resonance imaging and high-density mapping. J Am Coll Cardiol 62(9):802–812 7. Kanazawa H, Yamabe H, Enomoto K et al (2014) Importance of pericardial fat in the formation of complex fractionated atrial electrogram region in atrial fibrillation. Int J Cardiol 174(3): 557–564 8. Kopecky SL, Gersh BJ, McGoon MD et al (1987) The natural history of lone atrial fibrillation. A population-based study over three decades. N Engl J Med 317(11):669–674 9. Wijffels MC, Kirchhof CJ, Dorland R et al (1995) Atrial fibrillation begets atrial fibrillation. A study in awake chronically instrumented goats. Circulation 92(7):1954–1968 10. Tanaka K, Zlochiver S, Vikstrom KL et al (2007) Spatial distribution of fibrosis governs fibrillation wave dynamics in the posterior left atrium during heart failure. Circ Res 101(8): 839–847 11. Konings KT, Smeets JL, Penn OC et al (1997) Configuration of unipolar atrial electrograms during electrically induced atrial fibrillation in humans. Circulation 95(5):1231–1241 12. Lin J, Scherlag BJ, Zhou J et al (2007) Autonomic mechanism to explain complex fractionated atrial electrograms (CFAE). J Cardiovasc Electrophysiol 18(11):1197–1205 13. Kim RJ, Wu E, Rafael A et al (2000) The use of contrast-enhanced magnetic resonance imaging to identify reversible myocardial dysfunction. N Engl J Med 343(20):1445–1453 14. De Cobelli F, Pieroni M, Esposito A et al (2006) Delayed gadolinium-enhanced cardiac magnetic resonance in patients with
1199
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
chronic myocarditis presenting with heart failure or recurrent arrhythmias. J Am Coll Cardiol 47(8):1649–1654 Flett AS, Hasleton J, Cook C et al (2011) Evaluation of techniques for the quantification of myocardial scar of differing etiology using cardiac magnetic resonance. JACC Cardiovasc Imaging 4(2):150–156 Hwang SH, Choi EY, Park CH et al (2014) Evaluation of extracellular volume fraction thresholds corresponding to myocardial late-gadolinium enhancement using cardiac magnetic resonance. Int J Cardiovasc Imaging 30(Suppl 2):137–144 Henriksson KM, Farahmand B, Johansson S et al (2010) Survival after stroke—the impact of CHADS2 score and atrial fibrillation. Int J Cardiol 141(1):18–23 Verma A, Wazni OM, Marrouche NF et al (2005) Pre-existent left atrial scarring in patients undergoing pulmonary vein antrum isolation: an independent predictor of procedural failure. J Am Coll Cardiol 45(2):285–292 Oakes RS, Badger TJ, Kholmovski EG et al (2009) Detection and quantification of left atrial structural remodeling with delayedenhancement magnetic resonance imaging in patients with atrial fibrillation. Circulation 119(13):1758–1767 Akkaya M, Higuchi K, Koopmann M et al (2013) Higher degree of left atrial structural remodeling in patients with atrial fibrillation and left ventricular systolic dysfunction. J Cardiovasc Electrophysiol 24(5):485–491 Badger TJ, Daccarett M, Akoum NW et al (2010) Evaluation of left atrial lesions after initial and repeat atrial fibrillation ablation: lessons learned from delayed-enhancement MRI in repeat ablation procedures. Circ Arrhythm Electrophysiol 3(3):249–259 Taclas JE, Nezafat R, Wylie JV et al (2010) Relationship between intended sites of RF ablation and post-procedural scar in AF patients, using late gadolinium enhancement cardiovascular magnetic resonance. Heart Rhythm 7(4):489–496 Malcolme-Lawes LC, Juli C, Karim R et al (2013) Automated analysis of atrial late gadolinium enhancement imaging that correlates with endocardial voltage and clinical outcomes: a 2-center study. Heart Rhythm 10(8):1184–1191 Malcolme-Lawes SH, Oh YW, Lee DI et al (2014) Evaluation of quantification methods for left arial late gadolinium enhancement based on different references in patients with atrial fibrillation. Int J Cardiovasc Imaging. doi:10.1007/s10554-014-0563-0 Schmidt M, Dorwarth U, Andresen D et al (2014) Cryoballoon versus RF ablation in paroxysmal atrial fibrillation: results from the German Ablation Registry. J Cardiovasc Electrophysiol 25(1):1–7 Ho SY, Anderson RH, Sanchez-Quintana D (2002) Atrial structure and fibres: morphologic bases of atrial conduction. Cardiovasc Res 54(2):325–336 Atienza F, Calvo D, Almendral J et al (2011) Mechanisms of fractionated electrograms formation in the posterior left atrium during paroxysmal atrial fibrillation in humans. J Am Coll Cardiol 57(9):1081–1092 Park JH, Park SW, Kim JY et al (2010) Characteristics of complex fractionated atrial electrogram in the electroanatomically remodeled left atrium of patients with atrial fibrillation. Circ J 74(8):1557–1563
123
ORIGINAL PAPER
Relation between left atrial wall composition by late gadolinium enhancement and complex fractionated atrial electrograms in patients with persistent atrial fibrillation: influence of non-fibrotic substrate in the left atrium Sung Ho Hwang1 • Yu-Whan Oh1 • Dae In Lee2 • Jaemin Shim2 • Sang-Weon Park2 Young-Hoon Kim2,3
•
Received: 14 March 2015 / Accepted: 6 May 2015 / Published online: 10 May 2015 Ó Springer Science+Business Media Dordrecht 2015
Abstract The complex fractioned atrial electrogram (CFAE) has been considered as the catheter ablation target of left atrium (LA) under persistent atrial fibrillation (PeAF). We evaluated the relation between the LA wall composition by late gadolinium enhancement cardiac magnetic resonance (LGE-CMR) and the CFAE in patients with PeAF. Forty-three patients underwent LGE-CMR and CFAE mapping before catheter ablation for PeAF. The LA wall substrates were classified into three: the fibrotic, intermediate, and normal substrates by using two thresholds of 2 standard deviation (2-SD) and 6-SD above the mean signal from the normal myocardium. For each of 12 preselected LA wall regions, the composition ratios (CRs) of fibrotic, indeterminate, and normal substrates were calculated as a percentage to the volume of LA wall region, and compared depending on the CFAE, respectively. The CR of normal substrate was significantly greater at the LA wall region with CFAE (52 ± 38 % vs. 20 ± 28 %, P \ 0.01) than without CFAE. In contrast, the LA wall region with CFAE showed significantly lower CRs of intermediate substrate (39 ± 34 % vs. 57 ± 31 %, P \ 0.01) and fibrotic substrate (7 ± 17 % vs. 21 ± 24 %, P \ 0.01) than did the LA wall region without CFAE, respectively. Thus, the high CR ([18 %) of normal
& Young-Hoon Kim [email protected] 1
Department of Radiology, Korea University Medical Center, Seoul, Republic of Korea
2
Division of Cardiology, Department of Internal Medicine, Korea University Medical Center, Seoul, Republic of Korea
3
Division of Cardiology, Department of Internal Medicine, Anam Hospital, Korea University, Anam-dong 5-ga, Seongbuk-gu, Seoul 136-705, South Korea
substrate predicted the CFAE at the corresponding LA wall region with 71 % sensitivity and 62 % specificity. In conclusion, the evaluation of LA wall normal substrate by LGE-CMR might be useful to predict the CFAE occurrence before catheter ablation of PeAF. Keywords Left atrium Atrial fibrillation Fibrosis Myocardium Magnetic resonance imaging
Introduction Catheter ablation for electrical isolation of arrhythmogenic substrate has gained wide acceptance for the management of refractory atrial fibrillation (AF) [1, 2]. The pulmonary veins (PVs) have been considered as the main drivers of persistent AF (PeAF), and the electrical isolation of PVs can help achieve restoration of sinus rhythm in patients with PeAF [3]. Beyond the PVs, the complex fractionated atrial electrogram (CFAE) has been proposed as an important left atrial (LA) phenomenon maintaining PeAF. Catheter ablation strategies targeting the LA wall associated with the CFAE in addition to isolation of PV have been applied to improve the maintenance of sinus rhythm in patients with refractory PeAF [4, 5]. The CFAE is defined as complex electrograms with two or more deflections, continuous activity, and cycle length (CL) B120 ms [5]. Delineation of transient CFAE needs an invasive catheter-based electrophysiologic study. To minimize the invasive procedures, prior studies have evaluated the CFAE by using noninvasive methods such as computed tomography and magnetic resonance systems [6, 7]. Several studies about the LA wall substrates have suggested that PeAF is associated with progressive electrical and structural LA wall remodeling [8–10]. It has also
123
1192
been known that the summation of electrograms from overlying layers of normal myocardial fibers, autonomic nerves, and rotors might be related to the CFAE [11, 12]. Thus, we believe that a detailed assessment of LA wall composition may improve the prediction of transient CFAE occurrence. Late gadolinium enhancement cardiac magnetic resonance (LGE-CMR) has been considered a gold standard modality for differentiation between the fibrosis and normal myocardium in various cardiovascular diseases [13, 14]. Jadidi et al. [6] reported that the patchy LA fibrosis detected with LGE-CMR is inversely associated with the CFAE occurrence. So, we could surmise that the non-fibrotic substrates might also be associated with the CFAE occurrence in the LA wall. In the present study, we evaluated the LA wall composition of various substrates determined by LGE-CMR, depending on the CFAE map results. Furthermore, we tried to determine the feasibility of LGE-CMR in the prediction of CFAE occurrence in patients with PeAF.
Methods Study population Between July 2013 and July 2014, 65 patients sequentially underwent preprocedural cardiac magnetic resonance (CMR) examination for the evaluation of the LA and PVs, and CFAE mapping before catheter ablation for PeAF at the Korea University Anam Hospital. The patient with PeAF had an atrial fibrillation episode that lasted longer than 7 days or required termination by cardioversion without catheter ablation. In this study, the exclusion criteria were contraindication to CMR examination (e.g., severe claustrophobia, automatic implantable defibrillators, or pacemakers) and unacceptable quality of CMR images Eventually, of all selected 65 patients, 22 (33.8 %) were excluded due to the unacceptable quality of LGE-CMR images in the quantification of LA wall composition. A total of 43 patients (35 men; mean age, 54.4 ± 9.6 years) were enrolled into the present study. During CMR data acquisition, the mean heart rate of all enrolled patients was 69.2 ± 15.2 beats/min. No patients showed AF attack during the CMR examination. This retrospective study was approved by the institutional review board. All patients provided written informed consent for their participation. CMR protocol Before catheter ablation, CMR examination was performed by using a 3-T magnetic resonance system (Achieva; Philips Medical Systems, Best, the Netherlands) with a
123
Int J Cardiovasc Imaging (2015) 31:1191–1199
32-element phased-array cardiac coil. First, a three-dimensional (3D) spoiled gradient-echo contrast-enhanced timing robust angiography (CENTRA) was performed to define the cardiac anatomy of the LA and the PVs after the injection of 0.2 mmol/kg gadolinium contrast (Dotarem; Guerbet, S.A., Villepinte, France). Then, high spatial resolution LGE-CMR images were acquired approximately 15–20 min after the administration of the gadolinium contrast by using a 3D inversion recovery-prepared, respiration-navigated, electrocardiography (ECG) gated, gradient-echo pulse sequence after the Look–Locker sequence to identify the optimal nulling time for the normal left ventricular (LV) myocardium. Typical data acquisition parameters for LGE-CMR were free breathing by using a respiratory navigator with a 6-mm acceptance window; a transverse imaging volume with a voxel size of 1.5 9 1.5 9 1.5 mm; TR/TE of 4.7/1.4 ms; TI of 230–270 ms; flip angle of 25°; bandwidth of 127 Hz/pixel; 1 R wave to R wave (RR) interval between inversion pulses; phase encoding in the right–left direction; and parallel imaging with the sensitivity encoding (SENSE) technique with R = 2. The LGE-CMR images were acquired within a window of 100–150 ms for each RR wave interval. Respiratory navigator inflow artifacts were reduced by lowering the navigator rescales and positioning the navigator away from the right-sided PV. Depending on the successful leading navigator placement, free-breathing images aimed for a navigator efficiency of [30 %. Then, all LGE-CMR images were transferred into the workstation for image analysis. Analysis of LGE-CMR images All LGE-CMR images were analyzed independently by two experienced radiologists (with 7 and 20 years of experience in cardiac image interpretation) blinded to the patients’ clinical and electrophysiological data, by using a commercial software workstation (Terarecon iNtuition; TeraRecon, Foster City, CA, USA). Figure 1 shows the 3D reconstruction process of LGECMR images. On transverse or coronal LGE-CMR images, the epicardial and endocardial borders of the LA wall were semiautomatically contoured to select the entire LA wall and reconstruct a 3D LA model. In the quantitative analysis of the selected LA wall, we used signal threshold methods involving the manual delineation of regions of interest (ROI) that included, to the greatest possible extent, the normal left ventricular (LV) wall of nulled signal intensity on LGE-CMR image. On the basis of the mean and standard deviation (SD) values from ROI, the 6-SD and 2-SD thresholds above the mean signal in the normal LV wall were calculated [15, 16] (Fig. 1). By using the 6-SD and 2-SD thresholds, all LA wall substrates were classified into
Int J Cardiovasc Imaging (2015) 31:1191–1199
1193
Fig. 1 Determination of the normal, intermediate, and fibrotic substrates in the left atrial (LA) wall by using LGE-CMR. The 2-standard deviation (SD) and 6-SD threshold algorithms are established from the left ventricular myocardium of nulled signal intensity on the LGE-CMR image (a). The epicardial and endocardial borders of LA wall (green color) on LGE-CMR images (b, c) are semiautomatically selected for the reconstruction of the 3D LA wall
model (d). In the characterization of each of different LA wall substrates, the fibrotic substrates with signal intensities (green color) equal to or greater than the 6-SD threshold are selected in the posterior LA wall (white arrow) on transverse LGE-CMR image (e). In addition, the normal substrates with signal intensities (green color) less than the 2-SD threshold are selected in the LA roof (black arrow) on coronal LGE-CMR image (f)
three: (1) the fibrotic substrate with signal with [6-SD threshold; (2) the intermediate substrate with a signal of C2-SD threshold and \6-SD threshold; and (3) the normal substrate with signal of \2-SD threshold. Furthermore, different color look-up table masks were applied to the yellow for the fibrotic substrate, the gray for the indeterminate substrate, and the blue for the normal substrate on the 3D LA model for better delineation of LA wall composition (Figs. 1, 2).
system; St. Jude Medical, St. Paul, MN, USA) was generated. Bipolar intracardiac electrograms were filtered between 30 and 300 Hz. Intracardiac recordings were simultaneously recorded with CARTO and a computerized multichannel recording system (EP Med Systems Inc., Mt. Arlington, NJ, USA). Tachycardia and CL were monitored and recorded from the mapping catheter for the LA wall. The CFAE was defined as atrial electrograms with CL B120 ms [5]. For a CFAE map, as previously described, contact bipolar electrograms were obtained during an atrial fibrillation episode lasting[3 min; they were measured for more than 6 s at each LA wall site and were approximately equally distributed within the LA wall. The mean CL of the local electrogram was projected onto the LA wall shell with a color-coded display (Fig. 3). Additionally, the LA wall pixels of the CFAE were coded as red or white in the CFAE map.
CFAE mapping For recording and stimulation, a duodecapolar catheter (St. Jude Medical, St. Paul, MN, USA) and a decapolar catheter (Bard Electrophysiology, Lowell, MA, USA) were placed in the coronary sinus and in the low right atrium (RA) and high RA through the left femoral vein, respectively. A quadripolar catheter was also placed in the superior vena cava. After double transseptal puncture, the patients were administered with anticoagulants such as intravenous heparin to maintain an activated clotting time of between 300 and 400 s. Then, a 10-bipole or 20-bipole double-spiral circular mapping catheter (Lasso; Biosense Webster, Diamond Bar, CA, USA) was used for electrophysiologic mapping. Then, the CMR angiography merged electrophysiologic map (Ensite NavX
Comparison between LGE-CMR and CFAE map The entire LA wall was subdivided into 12 preselected LA wall regions: 6 on the posterior wall, and 6 on the anterior wall and septum (Fig. 4). For each of 12 LA wall regions, a quantitative analysis was performed to correlate the LGECMR images and CFAE map (Fig. 5). To determine the composition ratio (CR) of LA wall for each LA wall
123
1194
Int J Cardiovasc Imaging (2015) 31:1191–1199
Fig. 2 Three-dimensional (3D) reformatting and color coding of LA wall in basis of LGE-CMR for the characterization of left atrial (LA) wall composition. The normal substrate (blue), intermediate substrate (gray), and fibrotic substrate (red) can be simultaneously separated from the 3D LA wall model, transverse and coronal LGECMR images. Transverse LGE image shows a predominance of normal substrate in the anterior LA wall at the aortic valve level (arrow). Coronal LGE image shows a predominance of normal substrate in the LA roof (arrow)
Fig. 3 Anteroposterior view and left oblique view of the CFAE map. The LA wall pixels showing a mean cycle length (CL) of B120 ms are coded as red or white, presenting the CFAE occurrence
region, the summation of entire LA wall substrates was considered as the total volume of LA wall region. The mean total volume of every LA wall region was
123
3.1 ± 0.9 cm3. Furthermore, each CR of the normal substrate, the intermediate substrate, and the fibrotic substrate was also calculated as a percentage of the total volume of
Int J Cardiovasc Imaging (2015) 31:1191–1199
1195
Fig. 4 Schematic representation of the 12 preselected LA regions on the LGE-CMR image. On the posterior and anterior views of the 3D LA wall model, the 12 preselected LA wall regions include 6 regions at posterior LA wall (blue), 2 regions at LA septum (green), and 4 regions at anterior LA wall (yellow)
Fig. 5 Correlation between the LGE-CMR image and the CFAE map. The 3D LA wall model by LGE-CMR shows the focused distribution of the normal substrate (blue) in the anterior LA wall (white arrow) and posterior LA wall (black arrow), that well corresponds to the CFAE (red) with a cycle length of B120 ms on the CFAE map. In contrast, the LA septum with dense and large fibrotic substrates (yellow) on 3D LA wall model shows no CFAE
the LA wall at the corresponding LA wall region, respectively. In the analysis of LGE-CMR by the 2 investigators, intraobserver and intrerobserver reproducibility for the volume measurement of each LA wall substrates were calculated and found to be reliable (intraclass correlation coefficient [0.7 for all volume measurements). In the
evaluation of the CFAE within the corresponding LA region, the visual extent of the CFAE was evaluated by two blinded reviewers (two experts in electrophysiologic study) with consensus. Depending on the visual presence of the red or white-coded CFAE, every LA wall region was classified into two: (1) the CFAE region (with presence of
123
1196
Int J Cardiovasc Imaging (2015) 31:1191–1199
the white or red pixels in the corresponding LA wall region) and (2) the non-CFAE region (without the white or red pixels in the corresponding LA wall region). Statistical analysis Statistical analysis was performed by using SPSS software (version 19.0; IBM, Somers, NY, USA). All continuous data were expressed as mean ± SD, and all categorical data were presented as absolute values and percentages. Continuous variables were compared by using Student’s ttest depending on the CFAE map results such as the CFAE and non-CFAE at the preselected 12 LA wall regions. Receiver operating characteristic (ROC) curves were used to determine the accuracy of a variable in predicting the CFAE. A P value of \0.05 was considered statistically significant.
Results Table 1 shows the baseline clinical characteristics of the 43 enrolled patients. All patients had no diagnosis of coronary artery disease and cardiomyopathy. The mean LV ejection fraction was 54.2 ± 7.1 %. When considering the clinical history (e.g., hypertension, diabetes, heart failure, and stroke), the mean CHADS2 score [17] was 0.4 ± 0.8. Table 2 summarizes the frequency of CFAE and the CRs of three different LA wall substrates by LGE-CMR at every LA wall region. A high frequency ([50 %) of CFAE was noted in the posterior region (n = 22, 51 %), the superior septum region (n = 26, 60 %), the anterior roof region (n = 26, 60 %), and the appendage region (n = 26, 60 %). Additionally, the superior septum region, anterior roof region, and appendage region were located in the septum and anterior wall of the LA, which showed high mean CR of normal substrates [50 %. In contrast, the mean CR of fibrotic substrate was \5 % at the LA wall regions related to a high frequency ([50 %) of CFAE.
Table 1 Patients’ demographics (N = 43) Male
35 (81.3)
Age (years)
54.4 ± 9.6
Old age ([75 years)
0 (0)
Congestive heart failure
4 (9.2)
Hypertension
11 (25.5)
Diabetes mellitus
2 (4.6)
Stroke
2 (4.6)
CHADS2 score
0.4 ± 0.8
Data are presented as number (percentage) or mean ± standard deviation
123
When a total of 516 LA wall regions were evaluated, 152 (29 %) were assigned into the CFAE region. The mean CR of the normal substrate at the CFAE region was significantly greater than that at the non-CFAE region (52 ± 38 % vs. 20 ± 28 %, P \ 0.01) (Fig. 6). In contrast, the mean CR of the intermediate substrate at the CFAE region was significantly lower than at the non-CFAE region (39 ± 34 % vs. 57 ± 31 %, P \ 0.01) (Fig. 6). In addition, the mean CR of the fibrotic substrate at the CFAE region was also significantly lower than that at the nonCFAE region (7 ± 17 % vs. 21 ± 24 %, P \ 0.01) (Fig. 6). To establish the feasibility of the prediction of the CFAE region depending on the CR of each LA wall substrate by LGE-CMR, the sensitivity and specificity of various cutoffs related to the CR of each LA wall substrate were evaluated by using the ROC curve (Fig. 6). The areas under the ROC curves (AUCs) were 0.71 (P \ 0.01) for the CR of the normal substrate, 0.34 (P \ 0.01) for the CR of the intermediate substrate, and 0.31 (P \ 0.01) for the CR of the fibrotic substrate. When the CR of normal substrate was [18 % of the total volume of the LA wall at the preselected LA wall region, we could predict the CFAE occurrence at the corresponding LA wall region with a sensitivity of 71 % and a specificity of 62 % (Fig. 7).
Discussion We evaluated the LA wall composition by using LGECMR depending on the CFAE occurrence at each of the 12 preselected LA wall regions covering the entire LA wall in patients with PeAF. Among the 12 preselected LA regions, the posterior, superior septum, appendage, and anterior roof regions showed a high probability of CFAE occurrence. In the comparison of LA wall composition between the CFAE region and the non-CFAE region, the CFAE region showed a greater CR of the normal substrate, but lower CR of the fibrotic and intermediate substrates than did the non-CFAE region. Furthermore, we found that the high CR ([18 % of total LA wall volume at the preselected LA wall region) of the normal zone might be associated with the potential of CFAE at the corresponding LA wall region in patients with PeAF. LGE-CMR has been emerging as a powerful technique for identifying the hyperenhancement of fibrosis even in the thin LA wall [18–20]. Generally, patients with more remodeled LA wall have more extensive LA fibrosis than those with less remodeled LA wall [20, 21]. An appreciable correlation between LA fibrosis and low voltage (\0.5 mV) has been demonstrated [21, 22]. MalcolmeLawes et al. [23] revealed that the degree of hyperenhancement on LGE-CMR image correlates with the low tissue voltage representing the degree of fibrosis or injury.
Int J Cardiovasc Imaging (2015) 31:1191–1199 Table 2 Mean composition ratio (%) of LA wall substrate and the frequency of CFAE for each preselected LA wall region in 43 patients with persistent atrial fibrillation
1197
Normal
Intermediate
Fibrotic
CFAE occurrence
Posterior wall of the left atrium 1. Left superior PV region
12 ± 10
53 ± 21
36 ± 18
2 (5)
2. Posterior roof region
16 ± 21
78 ± 22
6 ± 11
13 (30)
3±5
76 ± 26
21 ± 26
2 (5)
4. Left inferior PV region
3. Right superior PV region
13 ± 9
70 ± 20
17 ± 18
2 (5)
5. Posterior region
49 ± 25
48 ± 24
3±6
22 (51)
6. Right inferior PV region
10 ± 8
69 ± 24
21 ± 21
2 (5)
Septum and anterior wall of the left atrium 7. Superior septum region
66 ± 32
34 ± 31
1±2
26 (60)
8. Anterior roof region
85 ± 9
12 ± 9
2±4
26 (60) 26 (60)
9. Appendage region
51 ± 44
48 ± 43
1±3
10. Inferior septum region
11 ± 13
29 ± 14
60 ± 6
4 (9)
11. Anterior region
11 ± 24
72 ± 33
17 ± 25
11 (26)
12. Mitral valve region
36 ± 40
40 ± 34
24 ± 27
17 (40)
Numbers in parentheses are percentages PV pulmonary vein
Fig. 6 Comparison of LA wall composition depending on the CFAE map result. There were significant differences in the composition ratio of each LA wall substrate (e.g., fibrotic substrate, intermediate substrate, and normal substrate) between the CFAE and the non-CFAE
By using LGE-CMR for tissue characterization, the ‘‘normal substrate’’ of baseline low signal intensity should first be defined because the fibrotic substrate and the intermediate substrate of hyperenhancement on LGE-CMR are identified on the basis of the relative signal difference from the ‘‘normal substrate’’ [13]. Signal intensity at 2-SD above the mean reference myocardium is the minimum threshold to differentiate between LGE and normal myocardium [16]. Despite the improved resolution of LGE-CMR image, the thin LA wall itself may sometimes disrupt the accurate segmentation of ‘‘normal substrate’’ on LGE-CMR images. Therefore, prior studies have used the high signal threshold based on references such as LA blood and LV wall for tissue characterization [23, 24]. Furthermore, the 6-SD
signal threshold of LGE can be feasible to define the LA wall corresponding to the low-voltage tissue on the electroanatomic map [24]. In the present study, the extreme signal thresholds by using the 2- SD and 6-SD thresholds based on the normal LV myocardium of nulled signal intensity were applied to define both the normal substrate and the fibrotic substrate. We believe that the classification of LA wall composition (e.g., normal, intermediate, and fibrotic substrates) according to the 2-SD and 6-SD threshold methods is somewhat applicable for the reproducible evaluation of the thin LA wall. The CFAE are not symmetrically located but can be predictably found in the certain LA wall regions, including the interatrial septum, posterior wall, anterior wall, roof,
123
1198
Fig. 7 Prediction of the CFAE in terms of the fibrotic, intermediate, and normal substrates by using the receiver operating characteristic (ROC) curve. At the preselected LA wall region, the volume of the normal substrate determined with LGE-CMR showed an area under the ROC curve (AUC) of 0.71. In addition, the high volume ([18 %) of normal substrate at the LA wall region predicted the CFAE occurrence with a sensitivity of 71 % and a specificity of 62 %
appendage base, and coronary sinus [5, 25]. In our study results, we also found that the LA wall regions frequently associated with the CFAE include the posterior, superior septum, appendage, and anterior roof regions of LA wall. In basis of our study results, the high frequency of CFAE in the septum and anterior wall of the LA was considerably compatible with the results from previous studies [5, 25]. In the aspect of anatomy, the anterior roof region and the appendage region include the overlapping layers of myocardial fibers such Bachmann’s bundle [11]. Additionally, the superior septum region where myocardial fibers from the left atrium and right atrium have specific muscular bundles has a high potential of CFAE [26]. Eventually, the ‘‘interatrial conduction’’ of muscular bundles or myocardial fiber may create CFAE with conduction abnormalities in the anterior roof, appendage, and superior septum regions. Thus, in the present study, we suggest that the ‘‘normal’’ substrate by using LGE-CMR might represent the presence of myocardial fiber, probably associated with the CFAE in the LA wall. Furthermore, the identification of LA wall composition by LGE before catheter ablation may provide an opportunity to characterize the PeAF pathophysiology. On the basis of our study results, the additional CFAE study should be recommended in PeAF patients with a high composition ratio of normal substrate by LGE especially in the anterior roof, appendage, and superior
123
Int J Cardiovasc Imaging (2015) 31:1191–1199
septum regions of LA wall. Substrate modification strategies for the CFAE have become increasingly used as part of ablation procedures in patients with PeAF [2, 5]. Although the mechanism of the CFAE is still unclear, the CFAE has been considered as the dyssynchronous activation of separate cell groups at pivot points, or the repetitive activations of the atrial fibrillation driver or local reentry circuit [27]. Nademanee et al. [5] reported that patients with PeAF have greater numbers and locations of CFAE in the LA wall than those with paroxysmal AF. In contrast, a prior study revealed that the LA wall area displaying the CFAE was smaller in long-lasting AF patients with a more electroanatomically remodeled LA wall than in those with less remodeled LA wall [28]. Jadidi et al. [6] demonstrated that the large fibrotic substrate detected with LGE-CMR is inversely associated with the CFAE occurrence in the LA wall. When we evaluated the various LA wall substrates of the low signal intensity or hyperenhancement by LGECMR, we also found that the fibrotic substrate of definite LGE was not a predominant component at the CFAE region of LA wall. Additional research is necessary to determine whether the prediction of CFAE occurrence by LGE-CMR may help the achievement of long-term success in the catheter ablation of PeAF. This study had several limitations. First, the sample size was small, and did not include a normal control group without a history of AF. A large follow-up study will be needed to define the clinical significance of noninvasive prediction of CFAE by using LGE-CMR before catheter ablation for PeAF. Second, the LGE-CMR resolution is limited for the evaluation of the thin LA wall. In this study, thin LA wall with a wall thickness of 1 mm may not be delineated or may be represented as the intermediate substrate with the resolution of the currently used LGE-CMR technique. Nevertheless, the reproducibility of measurement of the LA wall composition by using LGE-CMR was statistically reliable. With the advances in the LGE-CMR technique, it would be interesting to quantify the various LA wall substrates for evaluation of LA wall composition. Last, detailed studies with CFAE maps will be necessary to compare the CFAE map result and the LA wall signal intensity directly in the corresponding pixel points, because the visual assessment is limited for the accurate registration between the LGE-CMR and CFAE map. In summary, this study result emphasizes that the assessment of LA wall composition by using LGE-CMR may be useful to understand the mechanisms of CFAE in patients with PeAF. The fibrotic or intermediate substrates showing gray or bright signal intensity on LGE-CMR image might not sufficiently indicate a risk potential of CFAE in the preselected LA wall region. In contrast, the normal substrate of nulled signal intensity on LGE-CMR image might be associated with the CFAE occurrence in the
Int J Cardiovasc Imaging (2015) 31:1191–1199
preselected LA wall region under PeAF. Thus, the quantification of LA wall composition by using LGE-CMR can be useful for the prediction of the CFAE pattern in patients who will undergo catheter ablation for PeAF. Conflict of interest
None.
References 1. Haissaguerre M, Sanders P, Hocini M et al (2005) Catheter ablation of long-lasting persistent atrial fibrillation: critical structures for termination. J Cardiovasc Electrophysiol 16(11):1125–1137 2. Oral H, Chugh A, Yoshida K et al (2009) A randomized assessment of the incremental role of ablation of complex fractionated atrial electrograms after antral pulmonary vein isolation for long-lasting persistent atrial fibrillation. J Am Coll Cardiol 53(9):782–789 3. Haissaguerre M, Jais P, Shah DC et al (1998) Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med 339(10):659–666 4. Jais P, Haissaguerre M, Shah DC et al (1996) Regional disparities of endocardial atrial activation in paroxysmal atrial fibrillation. Pacing Clin Electrophysiol 19(11 Pt 2):1998–2003 5. Nademanee K, McKenzie J, Kosar E et al (2004) A new approach for catheter ablation of atrial fibrillation: mapping of the electrophysiologic substrate. J Am Coll Cardiol 43(11):2044–2053 6. Jadidi AS, Cochet H, Shah AJ et al (2013) Inverse relationship between fractionated electrograms and atrial fibrosis in persistent atrial fibrillation: combined magnetic resonance imaging and high-density mapping. J Am Coll Cardiol 62(9):802–812 7. Kanazawa H, Yamabe H, Enomoto K et al (2014) Importance of pericardial fat in the formation of complex fractionated atrial electrogram region in atrial fibrillation. Int J Cardiol 174(3): 557–564 8. Kopecky SL, Gersh BJ, McGoon MD et al (1987) The natural history of lone atrial fibrillation. A population-based study over three decades. N Engl J Med 317(11):669–674 9. Wijffels MC, Kirchhof CJ, Dorland R et al (1995) Atrial fibrillation begets atrial fibrillation. A study in awake chronically instrumented goats. Circulation 92(7):1954–1968 10. Tanaka K, Zlochiver S, Vikstrom KL et al (2007) Spatial distribution of fibrosis governs fibrillation wave dynamics in the posterior left atrium during heart failure. Circ Res 101(8): 839–847 11. Konings KT, Smeets JL, Penn OC et al (1997) Configuration of unipolar atrial electrograms during electrically induced atrial fibrillation in humans. Circulation 95(5):1231–1241 12. Lin J, Scherlag BJ, Zhou J et al (2007) Autonomic mechanism to explain complex fractionated atrial electrograms (CFAE). J Cardiovasc Electrophysiol 18(11):1197–1205 13. Kim RJ, Wu E, Rafael A et al (2000) The use of contrast-enhanced magnetic resonance imaging to identify reversible myocardial dysfunction. N Engl J Med 343(20):1445–1453 14. De Cobelli F, Pieroni M, Esposito A et al (2006) Delayed gadolinium-enhanced cardiac magnetic resonance in patients with
1199
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
chronic myocarditis presenting with heart failure or recurrent arrhythmias. J Am Coll Cardiol 47(8):1649–1654 Flett AS, Hasleton J, Cook C et al (2011) Evaluation of techniques for the quantification of myocardial scar of differing etiology using cardiac magnetic resonance. JACC Cardiovasc Imaging 4(2):150–156 Hwang SH, Choi EY, Park CH et al (2014) Evaluation of extracellular volume fraction thresholds corresponding to myocardial late-gadolinium enhancement using cardiac magnetic resonance. Int J Cardiovasc Imaging 30(Suppl 2):137–144 Henriksson KM, Farahmand B, Johansson S et al (2010) Survival after stroke—the impact of CHADS2 score and atrial fibrillation. Int J Cardiol 141(1):18–23 Verma A, Wazni OM, Marrouche NF et al (2005) Pre-existent left atrial scarring in patients undergoing pulmonary vein antrum isolation: an independent predictor of procedural failure. J Am Coll Cardiol 45(2):285–292 Oakes RS, Badger TJ, Kholmovski EG et al (2009) Detection and quantification of left atrial structural remodeling with delayedenhancement magnetic resonance imaging in patients with atrial fibrillation. Circulation 119(13):1758–1767 Akkaya M, Higuchi K, Koopmann M et al (2013) Higher degree of left atrial structural remodeling in patients with atrial fibrillation and left ventricular systolic dysfunction. J Cardiovasc Electrophysiol 24(5):485–491 Badger TJ, Daccarett M, Akoum NW et al (2010) Evaluation of left atrial lesions after initial and repeat atrial fibrillation ablation: lessons learned from delayed-enhancement MRI in repeat ablation procedures. Circ Arrhythm Electrophysiol 3(3):249–259 Taclas JE, Nezafat R, Wylie JV et al (2010) Relationship between intended sites of RF ablation and post-procedural scar in AF patients, using late gadolinium enhancement cardiovascular magnetic resonance. Heart Rhythm 7(4):489–496 Malcolme-Lawes LC, Juli C, Karim R et al (2013) Automated analysis of atrial late gadolinium enhancement imaging that correlates with endocardial voltage and clinical outcomes: a 2-center study. Heart Rhythm 10(8):1184–1191 Malcolme-Lawes SH, Oh YW, Lee DI et al (2014) Evaluation of quantification methods for left arial late gadolinium enhancement based on different references in patients with atrial fibrillation. Int J Cardiovasc Imaging. doi:10.1007/s10554-014-0563-0 Schmidt M, Dorwarth U, Andresen D et al (2014) Cryoballoon versus RF ablation in paroxysmal atrial fibrillation: results from the German Ablation Registry. J Cardiovasc Electrophysiol 25(1):1–7 Ho SY, Anderson RH, Sanchez-Quintana D (2002) Atrial structure and fibres: morphologic bases of atrial conduction. Cardiovasc Res 54(2):325–336 Atienza F, Calvo D, Almendral J et al (2011) Mechanisms of fractionated electrograms formation in the posterior left atrium during paroxysmal atrial fibrillation in humans. J Am Coll Cardiol 57(9):1081–1092 Park JH, Park SW, Kim JY et al (2010) Characteristics of complex fractionated atrial electrogram in the electroanatomically remodeled left atrium of patients with atrial fibrillation. Circ J 74(8):1557–1563
123
