Prediction of AF Ablation Results by Advanced Left Atrial Function Assessment Using Strain Analysis and Left Atrial Appendage Emptying Velocity

Abstract

Objectives

Atrial fibrillation (AF) is a prevalent arrhythmia with significant morbidity. Despite advancements in rhythm control strategies and ablation procedures, approximately 30 % of patients with paroxysmal AF experience recurrence, necessitating predictive tools for better patient stratification. This study evaluates the role of left atrial strain (LAS) and left atrial appendage emptying velocity (LAAeV) as predictors of recurrence, aiming to improve procedural outcomes and patient selection.

Methods

A prospective cohort of 32 patients with paroxysmal AF and structurally normal hearts, underwent either cryo- or radiofrequency ablation at a single tertiary center. Pre-ablation evaluations included LAS analysis via transthoracic echocardiography and LAAeV measurement via transesophageal echocardiography. Patients were followed for up to one year post-ablation, with recurrence defined as AF episodes lasting >30 seconds beyond a three-month blanking period. Statistical analyses assessed the predictive value of LAS and LAAeV, individually and in combination.

Results

AF recurrence occurred in 21.9 % of the cohort. Pre-ablation left atrial strain (LAS) values (global LAS ≥33.56 %, sensitivity 85.7 %, specificity 70 %, AUC = 81.4 %, P < 0.05) and left atrial appendage emptying velocity (LAAeV) (≥42.7 cm/s, sensitivity: 100 %, specificity: 87 %, AUC = 94 %, P < 0.001) were significantly associated with freedom from recurrence. The combination of LAS and LAAeV improved predictive accuracy to 100 % sensitivity and 94 % specificity (P value < 0.001). Post-ablation LAS showed inconsistent predictive value, with only apical-2 chamber global LAS achieving statistical significance (P = 0.002). Functional recovery of the left atrium post-ablation was minimal, suggesting limited reversibility of atrial remodeling.

Conclusion

Pre-ablation assessment of LAS and LAAeV provides robust predictors of recurrence in patients undergoing AF ablation. Incorporating these metrics into standard pre-procedural evaluation could optimize patient selection and improve ablation outcomes. Further studies are required to validate age-specific cutoff values and evaluate long-term implications.

1. Introduction

A trial fibrillation (AF) is the most common sustained cardiac arrhythmia, characterized by rapid and irregular electrical activity in the atria. It poses significant challenges in terms of its impact on patient health, healthcare systems, and economic burden.

The prevalence of AF increases with age, with the elderly population being particularly affected. It is estimated that approximately 2 %–4 % of the general population worldwide has AF, and this number is expected to rise as the population ages [1].

For a long time, there was a debate on which is the best management strategy for AF, rate or rhythm control. Growing evidence suggests that rhythm control is superior to rate control whenever feasible for both paroxysmal and persistent AF. It provides better quality of life and for patients with heart failure, rhythm control provides mortality benefit [2]. Recent studies as EAST AFNET (Early Treatment of Atrial Fibrillation for Stroke Prevention Trial–Atrial Fibrillation Network) suggested that early rhythm control is superior to rate control in composite end points of death, stroke and hospitalization regardless of ejection fraction (EF) [3].

Despite advances in management, rhythm control modalities are far from perfect, with best control ablation modalities resulting in around 30 % recurrence even in patients with paroxysmal AF. The mechanisms of recurrence are variable and reflect the complex and progressive nature of AF [4].

Left atrial myopathy is a newly recognized entity, involving clinical and subclinical structural and functional alterations of the left atrium. These changes are incorporated in the pathogenesis of atrial fibrillation and possibly in recurrence [5]. Detecting these changes can help modify our current practice and improve patients’ selection for different AF ablation procedures.

This study aims to assess predictors of recurrence, especially two strong predictors related to the atrial function: the left atrial strain (LAS) and the left atrial appendage emptying velocity (LAAeV). These parameters can be easily incorporated into the clinical practice, thus improving patients’ selection for procedures or guiding to extra sets of ablations in the first procedure according to the recurrence risk. It also can answer the question of when ‘early’ rhythm control is truly early. This is proposed to improve the overall outcomes for the patient and the healthcare system.

2. Methods

2.1. Population of the study

The study involved patients aged 18 or older with paroxysmal AF and structurally normal hearts, treated at Aswan Heart Center using catheter ablation (RF or CB) for pulmonary veins isolation (PVI). All patients signed an informed consent.

Routine investigations before AF ablation procedure included full laboratory investigations (labs) including complete blood count (CBC), renal and liver functions, coagulation profile and thyroid function tests, 12 lead electrocardiogram (ECG), review of the documented tachycardia ECG, ECG gated transthoracic echocardiography study, cardiac CT with contrast to delineate the pulmonary veins anatomy, and trans-esophageal echocardiography (TEE) 1–2 days pre-ablation to exclude left atrial appendage thrombus. All these investigations were saved on an electronic online archiving system for later review. All imaging tests were saved on the web based PAXERA achieving system (PaxeraHealth Corp™, Boston, MA.)

2.2. Selection of procedure type and procedural details

Cryo-AF ablation procedure was the standard procedure for first time ablation of paroxysmal AF. RF-ablation was reserved for patients with concomitant documented Flutter requiring extra lines as cavo-tricuspid isthmus (CTI) line or for patients with contraindications for Cryo-AF ablation as patients with contrast medium allergy.

2.3. Cryo ablation

Following transeptal puncture, The Arctic front™ Advance balloon and Achieve™ mapping catheter were introduced in each pulmonary vein. The balloon was inflated and after achieving good closure (confirmed by venography). Freezing was then done for a total of 240 seconds per vein, provided that PVs potential isolation occurred within 40 seconds. A delayed PVs potential isolation mandated stopping freezing and adjusting the balloon with better occlusion. A temperature of –60 °C mandated stopping application whenever achieved. During cryo ablation of right sided veins, the decapolar catheter was inserted into the SVCs to pace the phrenic nerve, starting from −10oc. Any alarming diaphragmatic contractions weakening mandated stopping the ablation.

2.4. RF ablation

3 D electro-anatomical mapping was done using Agilis deflectable sheath, high-density mapping catheter (PENTARAY™) and CARTO™ system. Bilateral WACA was done using 50 Watts RF ablation (smart touch™ catheter) with first pass isolation of all pulmonary veins, reaching target ablation index of 500/point for posterior side and 550/point for anterior side. At the end of the procedure and in sinus rhythm, entrance ± exit block of all veins was reconfirmed. Extra sets of ablations were performed if clinically indicated.

2.5. Post ablation protocol and follow up

All patients received class I or III anti-arrhythmic drugs for 3 months post-ablation (blanking period). In the 3 months post ablation follow up, all patients were assessed by history and by performing a 24 hour Holter for the recurrence of AF during the blanking period. In the same visit, ECG gated transthoracic echocardiography was performed to assess the changes in LA strain. In the 6 months and 1 year follow up, history and a 24 hour Holter monitor were used to assess the early recurrence post ablation. Definite recurrence was only considered in presence of a rhythm strip of more than 30 seconds documenting AF recurrence after the blanking period of 3 months.

2.6. Left atrial function assessment

During the routine preprocedural assessment, an ECG-gated trans-thoracic echocardiography was performed and later, offline analysis using special software to assess the left atrial strain.

LAS measurements were made offline by one trained operator, using vendor-independent software (TomTec imaging systems, version 1.4, Unterschleissheim, Germany), assessment was done in apical 4 chamber view and apical 2 chamber view and averages were reported for both 4- chamber and 2-chamber views using the onset of QRS as a zero reference for cardiac cycle. Strain measurements reflected the peak percentage change in length of the left atrial myocardium during the 3 phases of the cardiac cycle (systolic, early diastolic and late diastolic (atrial contraction) phases). The version available calculated the global left atrial longitudinal strain (global LAS). Other LA strain measurements (reservoir, conduit phases of LA function were not assessed independently. Delta global LAS was defined as the absolute change in the LAS from baseline to 3 months follow-up. The same software calculated left atrial (LA) volumes and EF (Fig. 1).

Fig. 1.

Example showing Tomtec interface and strain curves (88). LASr = left atrial strain reservoir function, which reflects the peak strain change during systolic phase of cardiac cycle. LAScd = left atrial strain conduit function, which reflects strain change in the early passive diastolic phase. LASct = left atrial strain contractile function, which reflects strain change during atrial contraction (late diastolic) phase.

2.7. Left atrial appendage emptying velocity measurement

During the routine TEE pre-intervention to exclude LAA thrombus or to guide trans-septal puncture, the LAAeV was assessed by pulsed wave Doppler over the LAA opening (Fig. 2).

Fig. 2.

Example showing LAAeV measurement. Here, the late diastolic emptying velocity wave measures 83.7 cm/s in a patient with no recurrence.

2.8. Statistical method

Data were coded and entered using the statistical package for the Social Sciences (SPSS) version 28 (IBM Corp., Armonk, NY, USA). Data was summarized using mean, standard deviation, median, minimum and maximum for quantitative variables and frequencies (number of cases) and relative frequencies (percentages) for categorical variables. Comparisons between groups were done using unpaired t test in normally distributed quantitative variables while non-parametric Mann–Whitney test was used for non-normally distributed quantitative variables. For comparing categorical data, Chi square (χ2) test was performed. Exact test was used instead when the expected frequency is less than 5. ROC curve was constructed with area under curve analysis to detect best cutoff value of significant parameters for detection of recurrence. Logistic regression was done to detect independent predictors of recurrence. P-values less than 0.05 were considered as statistically significant.

3. Results

We recruited a total of 45 patients with paroxysmal AF, with structurally normal heart, undergoing AF ablation between the start of 2020 till the end of 2022. Thirteen patients were excluded due to incomplete data, missed follow up visits or poor echocardiographic studies hindering proper analysis. Most of the patients had Cryo AF ablation (31/ 32) and one had 3D AF ablation (due to documented associated atrial flutter). Seven patients had documented recurrence after the blanking period (21.9 %). Three of the patients with recurrence had redo RF AF ablation. One patient had documented sustained short RP tachycardia in the 6-months follow up Holter. Electrophysiologic study was done and revealed AVNRT which was successfully ablated.

The mean age was 44.5 years. Males represented 71.9 % of population, The mean body mass index (BMI) was 31.18. The mean LV EF was 60.2 %. Regarding the comorbidities, obesity was prevalent in the population with a mean BMI of 31.18. Diabetes was present in 21.9 % of the patients, hypertension in 43.8 %, ischemic heart disease in 18.8 %. Smokers represented 25 % of the study population. None of the clinical parameters showed a significant association with AF recurrence postablation (see Table 1).

Table 1.

Population characteristics.

Age, mean (SD)	44.50 (11.85)
BMI, mean (SD)	31.18 (5.79)
LV EF, mean (SD)	60.16 (3.39)
Male, N (%)	23 71.9 %
DM, N (%)	7 (21.9 %)
HTN, N (%)	14 (43.8 %)
Smoker, N (%)	8 (25.0 %)
IHD, N (%)	6 (18.8 %)

The two studied parameters related to the atrial function; the left atrial strain and the left atrial appendage emptying velocity proved to be strong predictors for recurrences. Regarding the pre-ablation LAS, LAS measured in apical 2 views (LAS AP2), apical 4 views (LAS AP4) and when taking the average of the 2 views: The values were significantly higher in the no recurrence group compared to the recurrence group; LAS AP 4 (35.48 ± 7 Vs 23.9 ± 12, P value 0.045), LAS AP2 (35.95 ± 8.45 Vs 23.23 ± 10.45, P value 0.003) and for mean LAS (35.85 ± 7.1 Vs 23.54 ± 11.06, P value 0.026). The cutoff values for prediction of recurrence were as follows: For pre ablation LAS AP4 = 32.75 %, for LAS AP2 = 33.22 % and for their mean (mean LAS) = 35.56 % (Table 3). Regarding the LAS post ablation, only LAS measured in apical 2 could achieve statistical significance. Post ablation LAS AP2 was higher in the no recurrence group (36.37 ± 5.93 Vs 26.67 ± 8.05, P value 0.002) Tables 2 and 3.

Table 3.

Sensitivity, specificity and P values for the detected cut off values.

	Area Under the Curve	P value	95 % Confidence Interval		Cut off	Sensitivity %	Specificity %
			Lower Bound	Upper Bound
Pre-ablation LAS AP4	0.804	0.001	0.625	0.984	<32.75	71.4	78.3
Pre-ablation LAS AP2	0.818	<0.001	0.647	0.989	<33.22	85.7	63.6
Mean GLS pre	0.814	<0.001	0.638	0.991	<33.56	85.7	70
Pre-ablation mean LA EF	0.846	<0.001	0.675	1.016	<56.75	71.4	84
Post ablation LAS AP2	0.851	<0.001	0.675	1.027	<31.75	83.3	83.3
Post ablation mean LA EF	0.873	<0.001	0.744	1.003	61.425	100	68
LAAeV	0.942	<0.001	0.860	1.024	<42.7	100	87
Combined LAAeV and mean LAS pre-ablation	0.974	<0.001	0.916	1.032	0.3859496	100	94.7

Table 2.

Echo parameters and their relation to recurrence.

	Recurrence	No recurrence	P Value
LV EF (%)	61.43 ± 4.43	59.80 ± 3.06	0.268
Pre LAS AP4 (%)	23.91 ± 12.07	35.48 ± 7.20	0.045
Pre LAS AP2 (%)	23.23 ± 10.45	35.95 ± 8.45	0.003
Pre mean LAS (%)	23.57 ± 11.06	35.85 ± 7.10	0.026
Pre mean LA EF (%)	48.64 ± 13.85	63.11 ± 5.72	0.032
Post LAS AP 4 (%)	28.60 ± 10.98	34.80 ± 4.85	0.230
Post LAS AP2 (%)	26.67 ± 8.05	36.37 ± 5.93	0.002
Post mean LAS (%)	27.63 ± 9.29	35.39 ± 4.95	0.098
Post mean LA EF (%)	52.53 ± 10.40	64.12 ± 6.61	0.002
Delta LAS AP4 (%)	5.00 ± 7.63	−0.34 ± 4.55	0.077
Delta LAS AP 2 (%)	4.43 ± 8.29	0.75 ± 7.88	0.512
LAAeV (Cm/s)	36.78 ± 4.87	64.36 ± 19.68	<0.001

LAAeV was the strongest predictor of AF recurrence. The mean value was much higher in the no recurrence group with very high statistical significance. In a logistic regression model, it was an independent predictor of freedom from recurrence. A cutoff value of 42.7 cm/s was detected for recurrence prediction with a 100 % sensitivity and 87 % specificity with an AUC = 94 %, P value < 0.001 (Table 4). Combining LAAeV with the average pre-ablation LA GLS improved the prediction for recurrence to a model with 100 % sensitivity and 94 % specificity (Table 3).

Table 4.

Logistic regression for prediction of recurrence.

		P value	OR	95 % C.I.
				Lower	Upper
Recurrence	LAAeV	0.004	0.972	0.954	0.991

The mean LA EF, both pre and post ablation, were significantly higher in the no recurrence group. The cutoff values for the pre ablation LA EF was 56.75 % and for the post ablation EF was 61.46 % (Table 3).

4. Discussion

Our understanding of recurrence after AF ablation is still deficient, for instance some patients with pulmonary veins reconnection don’t experience recurrence. Many prediction models and parameters were tested to predict recurrence irrespective of the cause and mechanism of recurrence.

In our study, functional assessment of the left atrium is proved to be closely related to AF recurrence. Left atrial strain analysis measured in the apical 4 or apical 2 views and left atrial appendage emptying velocity were significantly correlated to the risk of the recurrence.

In this study, none of the traditional risk factors for cardiac diseases including age, sex, smoking, HTN, DM or presence of IHD could be correlated with higher risk of recurrence. Some of these factors are items in the CHADsVASc score which was tested to predict recurrence with average sensitivity and specificity, suggesting that not every point in the CHADsVAsc score has the same weight in predicting AF recurrence. Also, obesity as measured by BMI was not a predictor of recurrence in this study despite being prevalent in the population of study. The current evidence available for obesity suggests that being overweight (BMI >25 kg/m2) are associated with a greater risk of recurrent atrial arrhythmias after AF ablation (13 % increase for every 5 kg/m2 higher BMI). In the setting of comprehensive risk factor management, weight loss of ≥10 % in overweight and obese individuals with AF has been associated with reduced AF symptoms and AF burden in an RCT (aiming for BMI <27 kg/m2) [6]. A minimum of 10 % reduction is required to gain protective benefit in terms of decreased AF recurrence and average weight loss of 3–4 % had no effect on recurrence [7]. In this study, the effect of obesity was not evident as both groups had prevalent obesity/overweight characters.

Regarding the pre-ablation LAS, LAS measured in apical 2 views, apical 4 views and when taking the average of the 2 views, the values were significantly higher in the no recurrence group compared to the recurrence. The cutoff for prediction of recurrence were 32.75 % for pre ablation LAS AP4, 33.22 % for LAS AP2 and 35.56 % for their mean. In comparison to previous literature, a much lower cut off point of 24.3 % with a sensitivity of 97 % and specificity of 86 % was proposed [8]. This difference in cut off points can be explained by the different age groups enrolled in both cohorts. In the cohort with cutoff of 24.3 %, the mean age was 63 ± 7 years, in our cohort the mean age was 44 ± 12 years. It is a well-established fact that the normal LAS decreases with age [9]. This indicates that age specific cutoff points should be sought rather than a single cut off point.

On the other hand, for the post ablation LAS, only LAS measured in apical 2 could achieve statistical significance. The inconsistency of positive correlation of LAS post ablation with recurrence suggests that it’s the state of left atrium at time of ablation that is strongly related to recurrence. This supports the idea of early AF ablation to hinder AF progression due to clinical or subclinical left atrial changes which was proven beneficial in the recent EAST-AFNET trial. It was also found in this study that the changes in strain after AF ablation procedures by 3 months were not significant and couldn’t be correlated to recurrence. This suggests that the atrial changes that occur before ablation are not reversed by ablation. This goes in line with the recent results reported on 144 patients with follow-ups on atrial strain and its relation with recurrence, where the investigators found that there is a temporary worsening of LA strain function immediately after the procedure but this change gradually returns to baseline after 3 months [10].

Regarding the LA EF, the mean LA EF both pre and post ablation were significantly higher in the no recurrence group, emphasizing the same idea of better LA function correlating with no recurrence.

The best performing parameter was the LAAeV, In logistic regression, it was an independent predictor of freedom from recurrence. This can be explained by the fact that LAAeV is a surrogate for LA function and correlates well with the LAS. Another point is that it is much easier and simpler to measure, doesn’t require dedicated software and is less liable to errors in measurements. All this makes it a reliable and trusted method to assess the LA function and therefore a strong predictor of freedom from AF post recurrence. A cut off point of 42.7 could was detected for recurrence in this study. This goes in line with the present literature emphasizing the role of functional assessment of LAA. The reported cutoff point for recurrence in the literature after paroxysmal AF ablation is 38 cm/s [11]. It should be noted that the reported normal values for LAAeV ranged from 30 to 40 cm/ s. The proposed cutoff points in this study and in previous studies are in the high normal or slightly above normal range, indicating that recurrence in case of paroxysmal AF is related to subtle left atrial changes rather than marked gross abnormalities [12]. It will be of great value in future research to correlate values of LAAeV and atrial scars as detected by advanced MRI softwares. To the best of our knowledge, no age specific LAAeV normal ranges are available from population studies.

To improve the predictive power of the model, the LAAeV was combined with the average pre-ablation LAS. This model aims to assess the subtle LA changes by different methods. The predictive power of this model improved to 100 % sensitivity and 94 % specificity. P value < 0.001. This model performed better than all the available scores for prediction of the recurrence including a new score called VT-DHF score which combines multiple points including clinical, echo and ECG data [13].

5. Limitations

The sample size was not enough to fully assess the impact of LA strain post-ablation on AF recurrence. Also, there might be slight differences in measurements of LA GLS between different operators and using different software. However, this limitation is not present for measuring the LAAeV. The cut-off points for all these measurements should be adjusted to different age groups which adds another challenge. Finally, the performance of these parameters were not tested in the long term (>1 year), which needs further studies.

Another limitation is that the study was conducted during the 2 years of COVID-19 pandemic, a growing amount of data studied the effect of COVID-19 on LV and LA functions with cases with advanced LA dysfunction post COVID. Some recurrences in those years can be attributed to later COVID infections with newly developed atrial myopathies rather than progression of a previous disease. However, this specific limitation can be very hard to be bypassed in different clinical research after a pandemic.

6. Conclusion

LA function as assessed by LA GLS and by LAAeV are powerful predictors for AF recurrence post PVI. On the other hand, post-ablation LA GLS showed conflicting data based on the view of its measurement. No significant change in the left atrial function could be detected post-ablation or correlated with recurrence, suggesting that AF ablation procedures are not expected to reverse atrial functional changes, at least on the short follow up terms. The accurate assessment of LA function pre-ablation using LA GLS and LAAeV during the routine preparation of the procedure can predict patients with high risk of recurrence after PVI alone. These parameters are the most accurate parameters for prediction based on the results presented in this study with sensitivity of 100 % and specificity of 94 %.

Abbreviations

3D

Three-Dimensional

AF

Atrial Fibrillation

AUC

Area Under the Curve

AVNRT

Atrioventricular Nodal Reentrant Tachycardia

BMI

Body Mass Index

CB

Cryo-balloon

CBC

Complete Blood Count

CHADS-VASc

Congestive heart failure, Hypertension, Age ≥75 years, Diabetes mellitus, Stroke/Transient ischemic attack, Vascular disease, Age 65–74 years, Sex category

CTI

Cavo-tricuspid Isthmus

Delta LAS

Absolute Change in Left Atrial Strain from Baseline to Follow-Up

ECG

Electrocardiogram

EAST-AFNET

Early Treatment of Atrial Fibrillation for Stroke Prevention Trial—Atrial Fibrillation Network

EF

Ejection Fraction

GLS

Global Longitudinal Strain

LA

Left Atrium (or Left Atrial)

LAAeV

Left Atrial Appendage Emptying Velocity

LAS

Left Atrial Strain

LAS

AP2 Left Atrial Strain in Apical 2-Chamber View

LAS

AP4 Left Atrial Strain in Apical 4-Chamber View

LV

Left Ventricle

PVI

Pulmonary Vein Isolation

RCT

Randomized Controlled Trial

RF

Radiofrequency

SVC

Superior Vena Cava

TEE

Transesophageal Echocardiography

WACA

Wide Area Circumferential Ablation