Discussion
In this study, we established a joint model based on MR degree and left atrial appendage function assessed by preoperative echocardiogram in PAF patients, in order to predict the risk of recurrence after CPVA. The diagnostic efficiency of the model was validated. Atrial fibrillation has caused increasing burden on patients and the medical system. Catheter ablation has more advantages over drug therapy in the treatment of AF patients to maintain sinus rhythm[15], and has become the first-line treatment for patients with PAF[16]. The maintenance of sinus rhythm after ablation is related to reverse remodeling, which can reduce LA size and improve left ventricular (LV) function. However, because atrial fibrillation is more likely to recur after CPVA, early and late recurrences require multiple operations and drug adjustments, and the scar after ablation can aggravate the structural remodeling of the left atrium. Therefore, we need to find some preoperative non-invasive indicators to identify patients at high-risk of recurrence after ablation, which will facilitate the adjustment of individualized treatment plans and precision medicine.
Many diagnostic techniques have been proven to help predict the recurrence of atrial fibrillation, including blood tests, echocardiography, computed tomography, and electrophysiological examinations. In addition, researchers have established multiple prediction models through the multiple indicators, including APPLE, ATLAS, CAAP-AF scores, etc[17-19]. The detection echocardiography is cost effective, convenient, and widely accepted in daily workup[20]. Previous studies have shown that the recurrence after catheter ablation (CA) in AF patients is largely dependent on the changes in the structure and function of the left atrium[21, 22]. The PAF patients selected in this study were mostly in the early stages of AF, with normal or slightly enlarged left atrium inner diameter. Their changes in left atrial structure were not significant. Since the PAF patients selected in this study were mostly in the early stages of AF, it is necessary to further search for indicators that can reflect atrial function and remodeling. These indicators are easy to detect and can provide valuable information.
The left atrium plays a vital role in the normal operation of the mitral valve, and timely atrial contraction is very important for mitral valve closure[23]. The MR of patients with AF is mostly functional type I MR, which may be caused by dilation of the mitral valve annulus due to atrial remodeling. In addition, MR is also related to the patient’s age and the type of atrial fibrillation[24]. MR is easy to detect. The advances in equipment and clinical research have promoted more accurate methods for quantifying the degree of MR. Echocardiogram is particularly important in the evaluation of MR regurgitation. Other than the primary valvular disease and left ventricular insufficiency, 7.4% to 29% of AF patients have significant FMR[25-27], and 7.4% of the patients qualified for atrial fibrillation ablation have moderate or higher FMR. Recovering sinus rhythm can improve atrial function[24]. About 10.9% of PAF patients in this study had moderate or above FMR. Previous studies have shown that AF patients with MR are more likely to have recurrent AF after ablation. LZ et al. showed that the severity of MR in PAF patients was positively correlated with the incidence of recurrence after radiofrequency ablation[28]. Qiao et al. found that FMR was closely related to the matrix remodeling of the left atrium in PAF patients, and FMR was an independent risk factor for recurrence after radiofrequency ablation[29]. In this study, the risk of recurrence after radiofrequency in PAF patients gradually increased with the aggravation of mitral valve regurgitation. Previous research showed that the left atrial matrix remodeling , corresponding mitral valve annulus (MVA) expansion , mitral valve compensatory growth restriction, and decreased dynamic changes of the valve leaflets were important factors causing mitral valve regurgitation[26, 30]. After remodeling, the left atrium enlarges, the posterior leaflet annulus of the mitral valve shifts outward, and the anterior leaflet annulus shifts upwards by passive counterclockwise twisting; these changes lead to increased distance between the annulus papillary muscles, which limits the movement of mitral valve leaflet. In addition, the expansion of the mitral valve annulus is particularly obvious in the expansion of the posterior leaflet annulus[31]. The posterior leaflet of the mitral valve continues with the posterior wall of left atrium. The enlargement of left atrium causes the posterior leaflet of mitral valve to shift backward and downward. In addition, factors from left atrium, left ventricle, and annulus cause or aggravate the functional mitral valve regurgitation, which further promotes the enlargement of left atrium, leading to the occurrence and maintenance of atrial fibrillation[32]. Therefore, for PAF patients with insignificant left atrium enlargement, MR can be used as a predictor for LA matrix remodeling[29]. These studies all provide evidence for using MR as a potential marker of left atrial matrix remodeling in AF patients.
The left atrial appendage is an extremely important accessory structure of the left atrium. Its morphology and structure are more complex[33]. The loss of coordination between contraction and diastole of the left atrial appendage in AF patients results in slow internal blood flow. The transesophageal echocardiography before AF radiofrequency ablation allows the observation of patient’s left atrial appendage (LAA) hemodynamic changes from multiple angles, and yields the maximum emptying velocity of left atrial appendage. In addition, this study also used LAAV to rule out the effect of left atrial stunning on atrial function during the conversion of atrial fibrillation[34]. Therefore, LAAV plays an important role in clinical evaluation of left atrial appendage function. Melduni et al. found that, in patients with persistent AF after successful electrical cardioversion, the ones with reduced LAAV had increased risk of AF recurrence, stroke and death[35]. LAAV >40 cm/s could independently predict sinus rhythm persistence at 1 year[36]. In this study, the AUC of predicting AF recurrence based on LAAV alone was only 0.731, which may be related to the effects from heart function, heart rate, acquisition time and other factors on left atrial appendage function.
In this study, the degree of MR and LAAV were both independent predictors of atrial fibrillation recurrence after CPVA, but the diagnostic power of univariate indicator was still insufficient. Thus, we performed multivariate logistic regression on the observed indicators to establish a predictive model, which integrated clinical and echocardiographic parameters, and had more advantages over single-factor prediction. The calibration curve showed that the model had good prediction accuracy in both the modeling group and verification group. Many predicting models for recurrence after radiofrequency also showed similar results. He et al. found that a model constructed with LAAV could predict the risk of recurrence in PAF patients at 12 months after radiofrequency[37]. Yang et al. found that: the combined model integrating LA function and blood BNP level had good predictive value for the recurrence of early persistent AF after CPVA[38]. For the treatment of secondary MR, the management of primary disease is the most important; For severe cases, the MR guidelines recommend mitral valvuloplasty, mitral replacement, and percutaneous mitral valve repair. In this study, although the AF patients had moderate or higher MR before operation, their valve function was significantly improved after the recovery of sinus rhythm, and the degree of MR regurgitation was less than before. There was no patient with severe MR after radiofrequency.
This study still has limitations: (1) Although this study is a multi-center retrospective study, the number of cases included is small, and the small sample size may lead to selection bias; (2) All patients received 24-hour Holter monitoring during the follow-up, but they were not implanted with ring recorder, which may underestimate the recurrence rate of AF.
In summary, this study constructed and verified a model for predicting the recurrence of atrial fibrillation after CPVA in patients with PAF. The model took into account clinical and echocardiographic parameters and had good predictive performance. The model can identify high-risk patients with atrial fibrillation recurrence, and help doctors to optimize patient selection, inform patients of the recurrence risk, and design personalized treatment plans.