Article Text
Abstract
Imaging identifies patients with high-risk phenotype among the general population with atrial fibrillation, such as the presence of structural and valvular heart disease, which are both related to adverse outcome. Imaging is also potentially important for prediction of success of catheter ablation. Specifically, patients with larger left atrial size, reduced left atrial function and increased left atrial fibrosis content are more likely to experience atrial fibrillation recurrences after ablation. Routine and advanced echocardiographic imaging techniques and multi-detector row computed tomography and magnetic resonance imaging can provide detailed information. Currently, imaging techniques are not able to predict success on an individual basis, but it does permit identification of patients with high versus low risk of atrial fibrillation recurrence after ablation. Finally, imaging can be performed after ablation to demonstrate beneficial effects of restoration of sinus rhythm, including left atrial reverse remodelling and improvement in left atrial or ventricular function. All these issues are discussed in the current review.
Statistics from Altmetric.com
Introduction
Atrial fibrillation (AF) is the most frequent sustained arrhythmia and represents a major public health problem associated with significant morbidity and mortality. The prevalence of AF increases with age.1 In addition, AF is associated with a fivefold increased risk of stroke,w2 threefold risk of heart failurew3 and duplicates the risk of dementia and mortality.1 Structural heart disease (ie, ischaemic and valvular heart disease) and chronic diseases (ie, hypertension, hyperlipidaemia and chronic kidney disease) are associated with increased risk of AF.1 Stroke prevention and rhythm or rate control are the main objectives of AF treatment. Recent meta-analysis including 28 836 patients with AF suggests that rhythm control using catheter ablation techniques is more effective in reducing AF recurrences than antiarrhythmic drugs (odds ratio (OR) 5.87, 95% confidence interval (CI) 3.18 to −1085).2 However, identification of patients who will benefit from catheter ablation techniques remains challenging. Evaluation of arrhythmogenic substrate including left atrial (LA) size and function and tissue characterisation of the LA myocardium may help in predicting prognosis and outcome after catheter ablation. The current article provides an overview of the potential value of non-invasive cardiac imaging in patients with AF for risk stratification, for prediction of success of catheter ablation and to detect beneficial changes in cardiac function and structure after ablation.
Imaging for prognosis in the general population
A recent community-based study including 4618 individuals (mean age 73 years, 51% men) with new onset AF showed an increased risk of early (within 4 months after diagnosis; hazard ratio (HR) 9.62) and late mortality (after the first 4 months; HR 1.66) relative to age-matched and gender-matched general population (figure 1).3 Several imaging-based parameters have shown to be related with increased risk of all-cause and cardiovascular death and stroke in patients with AF (table 1).w4–w13 Presence of structural heart disease is one of the strongest prognostic parameters in patients with AF.w5–w7 Mitral stenosis and hypertrophic cardiomyopathy have been strongly related to increased risk of stroke in patients with AF.w5–w7 Accordingly, current guidelines coined the terms of ‘valvular AF’ and ‘AF with hypertrophic cardiomyopathy’.4 However, ‘non-valvular AF’ is more prevalent in the general population and for this entity several studies have demonstrated the role of imaging techniques for patient risk stratification (table 1).w4–w13 In the Atrial Fibrillation Follow-up Investigation of Rhythm Management trial,w12 LV hypertrophy, defined by an end-diastolic interventricular septum thickness >13 mm in men or >12 mm in women on 2D or M-mode echocardiography, was independently associated with all-cause mortality (HR 1.46, 95% CI 1.14 to −1.86, p=0.003) and stroke (HR 1.89, 95% CI 1.17 to 3.08, p=0.001). In addition, in the Belgrade Atrial Fibrillation Study,w11 a registry-based observational study including 842 patients with new onset AF and no structural heart disease, the risk of incident heart failure at follow-up was independently associated with LA diameter >40 mm (HR 1.8, 95% CI 1.1 to −2.8, p=0.018) and LVEF between 50% and −54% (HR 2.6, 95% CI 2.0 to −3.5, p<0.001). Additional studies have focused on the role of transoesophageal echocardiography (TEE) to predict the risk of stroke in patients with AF, since this imaging technique permits better evaluation of the LA appendage (LAA), main source of emboli in this population, and other parameters such as atherosclerosis of aortic valve and aorta.w9 w14–w17 Among 786 participants in the Stroke Prevention in Atrial Fibrillation-III trial,w9 TEE parameters were independently associated with an increased risk of thromboembolic events including the presence of LAA thrombi (relative risk 2.5, p=0.040), dense spontaneous echo contrast (relative risk 3.7, p<0.001), LAA peak flow velocities ≤20 cm/s on pulsed wave Doppler recordings (relative risk 1.7, p=0.008) and complex aortic plaque (relative risk 2.1, p<0.001). More recently with the use of 3-dimensional (3D) imaging techniques such as multi-detector row computed tomography (MDCT) or magnetic resonance imaging (MRI), Di Biase et alw8 showed that specific shapes of the LAA are associated with increased risk of thromboembolic events in patients with AF. While the role of 2-dimensional (2D) transthoracic echocardiography is pivotal in the risk stratification of patients with AF, the use of TEE or MDCT and MRI is commonly restricted to patients with additional cardiovascular risk factors (ie, prior stroke and diagnosis of AF) or undergoing specific therapies (electrical cardioversion or catheter ablation). However, the incremental value of the later imaging techniques over current clinical parameters included in the CHADS2 (Congestive heart failure, Hypertension, Age ≥75 years, Diabetes mellitus, prior Stroke or transient ischemic attack or thromboembolism) score has not been demonstrated.
Imaging prior to ablation
Various parameters derived from imaging may contribute to prediction of technical feasibility and success of catheter ablation. These include LA size, LA function, LA fibrosis and pulmonary vein anatomy (table 2)5–10 w18–w21 and can be assessed with various echocardiographic techniques, MRI and MDCT.
Assessment of the left atrium: size, structure and function
LA size. It has been demonstrated that the success rate of catheter ablation is less in the presence of severe LA dilatation.5 Transthoracic echocardiography is most often used in clinical routine and can provide linear and volumetric measurements of the left atrium. The LA anteroposterior diameter, obtained from the M-mode acquisition of the parasternal long-axis view (see online supplementary figure S1), has frequently been used in studies and clinical trials on AF.5 w22 However, this measure has been shown to underestimate the LA size.w23 w24 Volumetric measures, as assessed by the bi-plane Simpson's rule (see online supplementary figure S1), are currently recommended.w25 Still, 2D echocardiography underestimates significantly LA volume as compared with 3D imaging, which provides direct identification of the LA endocardial border in the 3D data set (see online supplementary figure S1). When 2D and 3D echocardiographies were compared with MRI (as the gold standard), maximum and minimum LA volumes assessed by 2D echocardiography were underestimated by 31±25 and 16±32 mL, respectively, whereas volumetric measurements were comparable between 3D echocardiography and MRI.11
MRI and MDCT provide high spatial resolution data and clear definitions of the endocardium improving the accuracy of 3D measurements of the LA. With MRI, normal values for LA volumes have been defined.12 A recent comparison between MDCT and MRI (n=34 patients) for LA volume assessment showed good correlations between the two techniques (r=0.85 and r=0.83 for maximum and minimum LA volumes).13 Furthermore, these imaging techniques permit evaluation of LA geometry, particularly, the sphericity index, which may characterise the LA remodelling pattern. In a single centre study including 106 patients undergoing AF ablation, LA sphericity index was significantly associated with AF recurrence at follow-up.14 MDCT however, is currently not the first choice technique for LA volume assessment due to the need for iodinated contrast and radiation exposure. It needs to be stated that in patients with AF, the beat-to-beat variability may influence on the accuracy of LA sizing.15
LA fibrosis. LA fibrosis is the hallmark of LA structural remodelling which is associated with new onset and perpetuation of AF. In addition, the extent of LA fibrosis is an independent predictor of success of catheter ablation. Direct assessment of LA fibrosis with non-invasive imaging techniques can be performed with echocardiographic-calibrated integrated backscatter imaging or with late gadolinium enhancement MRI (LGE-MRI).6 w26
Echocardiographic-calibrated integrated backscatter permits semiquantitative assessment of LA fibrosis by subtracting the ultrasound signal reflected by the LA myocardium from that of the pericardium (see online supplementary figure S2). Patients with AF recurrence after catheter ablation had lower magnitude of integrated backscatter of the LA (more fibrosis) compared with patients who remained in sinus rhythm (−13.9±4.0 vs −20.6±3.7 dB; p<0.001).6 Each 5 dB decrease in magnitude of calibrated integrated backscatter was associated with twofold higher risk of AF recurrence at follow-up.6 However, the low spatial resolution of this approach limits widespread use in clinical practice.
LGE-MRI provides high spatial resolution to visualise LA fibrosis. Gadolinium contrast agents remain trapped within the expanded extracellular matrix, which is visualised as hyperenhanced areas (bright). The group of Utah has provided cumulative data associating the extent of LA fibrosis with outcome after catheter ablation.8 w26 w27 This group has proposed a classification of LA structural remodelling based on the percentage of LA wall hyperenhancement (fibrosis): stage I is defined as <10% of the LA wall showing hyperenhancement, stage II as ≥10% and <20%, stage III as ≥20% and <30% and stage IV as ≥30% (figure 2).8 The multicentre prospective, observational Delayed-Enhancement MRI Determinant of Successful Radiofrequency Catheter Ablation of Atrial Fibrillation study included 260 patients with LGE-MRI data and complete follow-up and demonstrated that patients with extensive LA fibrosis (stage IV) had an unadjusted cumulative incidence of AF recurrence of 51% by day 325 as compared with 15.3% for patients with minimal fibrosis (stage I).8 For patients with stages II and III of LA fibrosis, the recurrence rates were 33% and 46%, respectively. Each 1% increment in LA fibrosis extent was associated with an HR of 1.06 (95% CI 1.03 to −1.08) for AF recurrence. However, these results should be interpreted with caution. Technical issues (accurate segmentation of the left atrium, application of hyperenhancement thresholds that accurately identify fibrous tissue) and reproducibility of data analysis need to be resolved to improve the performance of LGE-MRI to accurately select the patients who will benefit from catheter ablation.
LA function. In addition to LA size, LA function is also an important predictor of the success of catheter ablation.9 ,10 w28 The main function of the LA is to regulate LV filling by acting as a blood reservoir during LV systole, a conduit of blood from the pulmonary veins to the LV during early diastole and contributing to LV filling during the late diastole with its active booster pump function.
LA function can be assessed with conventional 2D and Doppler echocardiographic techniques, measuring from the apical four-chamber and two-chamber views, the maximum (before mitral valve opening), minimum (during mitral valve closure) and preatrial contraction (using the P-wave of the surface ECG as reference) volumes. The total, passive and active emptying fractions can be calculated which reflect the reservoir, conduit and booster pump functions, respectively. The use of 3D techniques has improved the accuracy of LA volume quantification and LA function assessment (see online supplementary figure S3).w18 w19 In 154 patients (36% with persistent AF) undergoing catheter ablation, 3D echocardiography-derived LA expansion index (reservoir function) was independently associated with AF recurrence (OR 0.99, 95% CI 0.980 to −0.998, p=0.019).9 Furthermore, Dodson et al7 recently showed that LA conduit function measured with MRI in 346 patients with AF was an important determinant of ablation success. Patients within the lowest LA conduit function quintile (mean 8.2% LA passive emptying function) had a fourfold increased risk of AF recurrence after 2 years of follow-up as compared with patients in the highest quintile (40.4% LA passive emptying function, indicating more preserved function).
These functions can be also measured using pulsed wave Doppler recordings of the transmitral flow and pulmonary vein flow. On transmitral flow velocities, the E-wave and A-wave reflect the LA conduit function and the booster pump function, respectively (see online supplementary figure S4). On pulmonary blood flow recordings, the peak pulmonary vein systolic (S) and diastolic (D) velocities reflect the LA reservoir and conduit function, respectively. Using intracardiac echocardiography, Verma et al w20 showed that patients who presented with AF recurrences after ablation had a significantly lower S-velocity compared with patients who remained in sinus rhythm (35.9±16.8 vs 45.9±22.4 cm/s, respectively; p=0.004) indicating more impaired LA reservoir function. Advances in MDCT data postprocessing software have permitted assessment of LV diastolic function and derivation of E-wave and A-wave which have been validated against Doppler echocardiographic techniques (figure 3).16 To obtain these measurements, MDCT data should be acquired across the entire cardiac cycle with ECG gating, which is associated with an increase in radiation.
LA reservoir function is highly determined by LA compliance and, therefore, indirect assessment of changes in the microstructure of the LA wall using deformation imaging techniques (tissue Doppler imaging (TDI) and speckle tracking-derived strain/strain rate) may provide a comprehensive assessment of the LA arrhythmogenic substrate (figure 4). Kuppahally et al17 demonstrated that patients with persistent AF had a higher LA fibrosis content (22±17% vs14±9%, p=0.04) and lower values of mid-lateral LA strain (35±16% vs 45±14%, p=0.003) compared with patients with paroxysmal AF. Furthermore, Schneider et al w28 showed that a value of strain rate >2.25/s measured at the septal or inferior LA walls or a value of strain >19.5% measured at the inferior LA wall was the best independent predictor of sinus rhythm maintenance after ablation. More recently, Morris et al10 demonstrated that LA strain (reservoir function) <18.8% and LA strain rate (booster pump function) >0.85/s were strongly associated with AF recurrences after ablation (OR 6.8 and 5.2; p=0.001 and p=0.003, respectively).
Finally, increased LA fibrosis may lead to a prolonged total atrial activation time and the creation of wavelets, macroreentries and localised sources that contribute to the AF substrate.18 ,19 w29 The time delay between the P-wave on the surface ECG and the mechanical activation of the LA, the so-called PA-TDI, can be measured using TDI techniques (see online supplementary figure S5). In a recent study including 213 patients undergoing AF ablation (22% persistent AF), PA-TDI was significantly longer in patients who presented with AF recurrences compared with patients who remained in sinus rhythm after the procedure (124±23 vs 146±20 ms, p<0.001).20 The accuracy of PA-TDI to predict AF recurrence was superior to that of maximum LA volume (area under the curve: 0.76 vs 0.56, respectively). Furthermore, combining non-invasive electrical LA mapping obtained from multiple-electrode vests applied to the patient's torso and LA anatomical data from MDCT, wave propagation patterns and their beat-to-beat changes can be evaluated using specific algorithms of signal filtering and phase mapping.18 ,19 Although not widely used, this novel technology refines the evaluation of the AF substrate and may help to personalise the ablation technique in each patient.
Pulmonary vein anatomy
MDCT and MRI provide 3D reconstructions of the LA and pulmonary vein anatomy that can be overlaid onto the 3D electroanatomical mapping providing a roadmap to guide the procedure and improve the success rate.w28–w30 The typical anatomy consists of four pulmonary veins with separate ostia (observed in 47%–88% of patients; see online supplementary figure S6).w21 w30 w31 Atypical pulmonary venous anatomy mainly includes an additional right middle pulmonary vein (described in 18%–25%) or common ostia for left pulmonary veins (observed in 13%–27%).w21 w30 w31 Although the association between atypical pulmonary venous anatomy and increased risk of AF recurrence is not consistent in the different series, it appears that the presence of a middle right pulmonary vein is more frequently observed among patients with AF recurrence.w30 w31 Moreover, larger dimensions of the pulmonary veins and their ostia have been associated with higher AF recurrence rate.w21
Imaging after ablation
Scar formation
LGE-MRI is currently the preferred imaging technique to evaluate LA scar formation after catheter ablation. Initial studies showed a good correlation between LA areas of LGE and the location of the ablation lines by electroanatomical mapping, suggesting the possibility of evaluating the completeness of circumferential pulmonary vein lesions and to predict AF recurrences after ablation and potentially guiding redo procedures.w32–w34 Despite the promising results, spatial resolution of LGE-MRI (around 2 mm) may be suboptimal for LA wall assessment (average thickness also approximately 2 mm) and, therefore, limits its application in the clinical practice.
Pulmonary vein stenosis
Pulmonary vein stenosis is an important complication after catheter ablation; the incidence of symptomatic patients has been estimated around 1%.w35 w36 Conventional invasive angiography or MDCT is recommended to diagnose this complication (figure 5). MRI is an alternative technique for the diagnosis and follow-up of pulmonary vein stenosis to limit radiation dose.
Changes in LA size and function
Significant LA reverse remodelling after successful catheter ablation has important prognostic implications and is considered a surrogate endpoint to measure success of ablation.21 w37 w38 Reduction in LA volume has been demonstrated after catheter ablation using 2D echocardiography and more advanced 3D techniques such as MRIw39 w40 and MDCT.w41 3D echocardiography has been used to evaluate the changes in LA volume and LA function (phasic volumetric changes) in patients undergoing catheter ablation.22 A significant reduction in LA volumes (maximum and minimum) and a significant improvement in LA active contraction and LA reservoir function were observed in patients with successful ablation (figure 6). The improvement in LA active function can be interpreted as a sign of improved atrial contractility, whereas the increase in LA reservoir function might reflect a reduction in LA stiffness and indirectly an improvement in LV systolic function.
Changes in LV systolic function
Successful catheter ablation may also have a favourable effect on LV systolic function. In patients with impaired LV function at baseline, a significant improvement in LVEF has been demonstrated after restoration of sinus rhythm following catheter ablation.23 Conversely, in patients with preserved LVEF, no change in LVEF was noted after successful catheter ablation.w42
Importantly, in patients with preserved LVEF, conventional methods may not be able to detect subtle LV dysfunction and its potential changes over time. 2D speckle-tracking strain analysis could be an alternative to detect subtle LV dysfunction despite preserved LVEF.24 w43 A recent study including 78 patients with a mean LVEF of 60%±7% who underwent catheter ablation showed a significant improvement in LV circumferential and longitudinal strain (and strain rate) only in the 54 patients who maintained sinus rhythm at follow-up (while LVEF remained unchanged); conversely, an impairment of these parameters was observed in the 24 patients with AF recurrences.24 These findings suggest a beneficial effect of restoration and maintenance of sinus rhythm on LV systolic function; the mechanism underlying these changes is unclear but might be mediated by a more efficient LV filling and by normalisation of the neurohormonal system.
Conclusion
Imaging identifies patients with high-risk phenotype among the general population with AF. In addition, imaging may become particularly useful for prediction of success of catheter ablation. At present, the imaging techniques are not able to predict success on an individual basis, but it does permit identification of patients with high versus low risk of AF recurrence after ablation. Randomised trials including imaging-based LA variables may establish the role of imaging in selection of patients for AF ablation techniques. Finally, imaging after ablation demonstrates LA reverse remodelling, and improvement in LA function and/or LV systolic function, which may positively affect long-term outcome.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
- Data supplement 2 - Online references
Footnotes
-
Contributors All authors contributed to this review.
-
Competing interests None.
-
Provenance and peer review Commissioned; externally peer reviewed.