External Cardioversion of Atrial Fibrillation Causes an Early Improvement of Cardiac Performance: A Longitudinal Strain Analysis Study.

INTRODUCTION: Atrial fibrillation (AF) is often associated with heart failure. Several studies have demonstrated that resumption of sinus rhythm (SR) improves cardiac output in the long-term. Aims of this study were to evaluate the acute variations of left ventricular (LV) performance, following successful external cardioversion (ECV) of persistent AF using longitudinal strain (LSt) analysis, and the influence of inflammation.
MATERIALS AND METHODS: We enrolled 48 patients with AF (age: 73 ± 12 years, men: 83.3%). A standard transthoracic echocardiographic evaluation was performed before the procedure and 6 h later; this included the analysis of LV endocardial peak LSt, a measure of myocardial deformation. In the last 32 patients, plasma concentration of interleukin-6 (IL-6) was also determined.
RESULTS: Restoration of SR led to the decrease of heart rate (HR) (74 ± 21 vs 64 ± 10 bpm, P < 0.001) and LV end-systolic volume (30 ± 16 vs 27 ± 17 mL/m2, P = 0.001), and to the increase of LV end-diastolic volume (LVEDV) (56 ± 20 vs 60 ± 21 mL/m2, P = 0.036) and ejection fraction (EF) (48 ± 10 vs 57 ± 11%, P < 0.001). Peak LSt improved in 43 (89.6%) patients (-12.9 ± 3.3 vs -18.0 ± 4.7%, P < 0.001). Multivariate analysis (R = 0.729, P < 0.001) showed that strain changes were directly correlated with basal HR and the appearance of atrial mechanical activity and inversely correlated with corrected thyroid dysfunction, LVEDV and the presence of a permanent pacemaker. Higher levels of IL-6 negatively affected LV performance improvement.
CONCLUSIONS: Effective ECV of AF determines a significant and fast improvement of LV performance, which is readily captured by LSt analysis. Inflammatory status may impact the response to SR restoration.

INTRODUCTION

Atrial fibrillation (AF) is the most frequent sustained arrhythmia observed in clinical practice with a tight association with age. In fact, its prevalence increases from 0.1% in subjects aged <55 years to 9% in those aged >80 years.[1] Many cases of AF are associated with heart failure (HF) where the arrhythmia is often detected in the 2 years preceding or following the diagnosis of HF.[12] Several physiopathological mechanisms may explain this association: In patients with systolic and diastolic HF left atrial volume is higher and left atrial emptying is lower than in controls,[3] enhancing the propensity to develop AF or flutter.[4] Interestingly, the presence of AF increases the risk of cardiovascular and noncardiovascular death in HF patients.[2]

External electrical cardioversion (ECV) of AF is the best method to restore sinus rhythm (SR) when the arrhythmia is present from >48 h; in fact, after that window the effectiveness of antiarrhythmic drugs is greatly reduced.[1] The European Society of Cardiology (ESC) and the American Heart Association/American College of Cardiology (AHA/ACC) current guidelines also recommend the use of ECV when a high ventricular rate does not respond promptly to pharmacological treatment in patients with ongoing myocardial ischemia, symptomatic hypotension or HF,[15] and in persistently symptomatic HF patients despite an adequate rate control.[1]

It is well-known that in the medium-long term AF patients, effectively treated with ECV, undergo a progressive, significant increment of cardiac output compared to baseline, with the improvement being correlated to the resumption of atrial mechanical activity.[67] However, some studies seem to suggest that, immediately after SR restoration, the degree of the change in atrial function inversely correlates with the length of the arrhythmia[8] and with the degree of atrial dilatation.[9] Moreover, there is evidence suggesting that, right after ECV, cardiac performance could worsen in some patients leading to the development of acute HF and pulmonary edema.[61011]

On this basis, the primary endpoint of this study was to verify if successful ECV of persistent AF was associated with acute favorable effects on left ventricular (LV) performance. For this purpose, we used longitudinal strain (LSt) analysis, a highly sensitive speckle tracking echocardiographic (STE) technique which allows the assessment of LV mechanics and their changes.[12] The secondary endpoint of the study was to measure the relation between inflammation and improvement of LV performance after SR restoration.

MATERIALS AND METHODS

Patients’ selection and clinical procedures

Between August 2010 and July 2011, all patients admitted to a day-hospital for ECV of persistent AF were enrolled in the study. The protocol was approved by the institutional review committee and complies with the Declaration of Helsinki. All subjects gave their informed consent. Clinical and instrumental assessments were carried out in accordance with the ACC/AHA recommendations for patients with AF.[13] Chronic obstructive pulmonary disease (COPD) and diabetes mellitus were diagnosed using the National Heart, Lung, and Blood Institute (NHLBI) and World Health Organization (WHO)[14] and the American Diabetes Association[15] criteria, respectively. The following cardiac diagnoses were considered as possible causes of AF:

Coronary heart disease in the absence of HF,

Stage C or D HF[16] of any etiology,

Significant valvular dysfunction in the absence of overt HF,[116]

Hypertension,[1] and

Lone AF.

ECV was performed after a 4-week period of effective oral anticoagulation.[1] Blood test and chest X-ray were obtained before ECV.

In the last 32 patients, preprocedure concentration of plasma interleukin (IL)-6 was determined using a commercially available enzyme-linked immunosorbent assay (ELISA) kit (Human Interleukin-6 UltraSensitive, Invitrogen Corporation, Life Technologies, Italy).[17]

Patients were excluded from the study for failure of ECV, immediate or early recurrence of AF,[1] or for an inadequate echocardiographic acoustic window.

ECV was performed according to the 2010 AHA Guidelines[18] as previously described.[19] Briefly, first shock was set at 130 J, if AF persisted, the energy was increased to 180 J. If the arrhythmia still persisted, two additional shocks were delivered after the infusion of an antiarrhythmic drug.[1] ECV was considered successful when the patient was discharged in SR.

Echocardiographic evaluation

A standard transthoracic echocardiogram was obtained immediately before the procedure (baseline evaluation) and at discharge (6 h after SR restoration), using a MyLab30Gold equipment (ESAOTE SPA, Florence, Italy).[20] We analyzed the peak values of LSt which measures the degree of LV deformation through the cardiac cycle and quantify regional myocardial function.[12] A specific software (XStrain™ - ESAOTE SpA) was employed for the analysis of digital content of cardiac images.[21] In detail, from an apical four-chamber view the endocardial border was marked with 13 tracking points; each tracking point is characterized by a specific speckle pattern on a gray scale, a sort of tissue fingerprint. Each tracking point was traced through the different phases of cardiac cycle using electrocardiogram (ECG) for synchronization. The measurement of spatial and temporal data related to the local displacement of each point, such as myocardial velocity and strain, correlates to parameters of LV contractility.[1222] Myocardial LSt is defined by the following formula:

LSt = (l - l0)/l0 = Δl/l0

where “l0” is the end-diastolic length of a segment and “l” is its end-systolic length. According to the above formula, LSt is a negative number and its absolute value is higher for better LV performance.[23] LSt data from six individual segments (i.e., basal, mid, and apical interventricular septum and apical, mid, and basal lateral wall), derived from three consecutive cycles, were averaged. The overall LSt was computed from the values of the individual segments [Figure 1].

Figure 1

Pre- (baseline) and post-external cardioversion (ECV) longitudinal strain analysis of left ventricular performance; the dotted line represents the overall value of the individual segments

Statistical analysis

Statistical analysis was carried out with IBM Statistical Package for the Social Sciences (SPSS) for Windows (version 18; Armonk, NY, USA). Continuous variables were expressed as mean ± SD; categorical variables were expressed as absolute numbers and percentages. Student's t-test and analysis of variance were used to compare continuous variables between two or more categories, respectively. If variables presented a non-normal distribution, the corresponding non-parametric tests — Mann-Whitney and Kruskal-Wallis — were employed. The relation between two continuous variables was studied with linear regression analysis. Changes of LSt and other variables, between AF and SR, were assessed with paired samples t-test or Wilcoxon test. The association between categorical variables was evaluated with chi-square test. Based on IL-6 concentrations, patients were stratified into three arbitrary groups: <0.16 pg/mL (lower limit for IL-6 detection; N = 14), 0.16-0.49 pg/mL (N = 12), and ≥0.50 pg/mL (N = 6). All variables associated with LSt changes in univariate analysis were entered into a multivariate linear regression model, using backward deletion. Among several collinear variables, only the one showing the highest degree of correlation with the dependent variable at univariate analysis was entered into the model. A P-value < 0.05 was considered statistically significant.

RESULTS

Study population

Fifty-eight patients underwent ECV; of those, 10 were excluded for inadequate acoustic window (N = 5) or failure of the procedure (N = 5). Hypertension was the most frequent CV disease; COPD, type 2 diabetes, and pharmacologically compensated dysthyroidism (hypothyroidism - N = 10, hyperthyroidism - N = 2) were the most common extracardiac diseases. Amiodarone was the most frequently used antiarrhythmic agent [Table 1]. Ten patients (20.8%) had a permanent pacemaker: Single- or dual-chamber device in eight and cardiac resynchronization therapy device in two cases. Left atrial enlargement (end-systolic diameter > 40 mm) was present in 81.0% of patients.

Table 1

Clinical and laboratory variables

Variables	(n = 48)	Range
Age (years)	73±12	36-89
Men (n, %)	40 (83.3)
Weight (kg)	78±15	56-113
BSA (kg/m2)	1.91±0.19	1.53-2.40
Smokers past-present (n, %)	20 (41.7)
CAD (n, %)	16 (33.3)
HF (n, %)	16 (33.3)
Dyslipidemia (n, %)	17 (35.4)
Hypertension (n, %)	38 (79.2)
Valvular disease (n, %)
Aortic	12 (25.0)
Regurgitation	7 (14.6)
Stenosis	4 (8.3)
Combined	1 (2.1)
Mitral	24 (50.0)
Regurgitation	23 (47.9)
Stenosis	0
Combined	1 (2.1)
CKD (n, %)	4 (8.3)
COPD (n, %)	7 (14.6)
CVD (n, %)	4 (8.3)
Diabetes (n, %)	6 (12.5)
Dysthyroidism (n, %)	12 (25.0)
Lone atrial fibrillation (n, %)	8 (16.7)
Therapy
ACE-I/ARB (n, %)	39 (81.3)
Beta-blockers (n, %)	30 (62.5)
Digoxin (n, %)	19 (39.6)
Loop diuretics (n, %)	17 (35.4)
K+ sparing agents (n, %)	4 (8.3)
Statins (n, %)	21 (43.8)
Oral anticoagulation (n, %)	45 (93.8)
ASA (n, %)	6 (12.5)
Class I antiarrhythmic drugs (n, %)	6 (12.5)
Amiodarone (n, %)	23 (47.9)
Class IV antiarrhythmic drugs (n, %)	1 (2.1)
BUN (mg/dL)	48±28	20-176
Creatinine (mg/dL)	1.0±0.4	0.6-2.5
K+ (mEq/L)	4.2±0.5	2.8-5.1
Hb (g/dL)	13.8±1.6	9.3-16.3
WBC (n·10-3/mm3)	6.5±1.5	3.7-9.4
Pro-BNP (mmol/L)	1967±2981	162-17541
ESR (mm/h)	16±12	2-50

BSA = Body surface area, CAD = Coronary heart disease, HF = Heart failure, CKD = Chronic kidney disease, COPD = Chronic obstructive pulmonary disease, CVD = Cerebrovascular disease, ACE-I = Angiotensin converting enzyme inhibitors, ARB = Angiotensin receptor blockers, ASA = Acetylsalicylic acid, BUN = Blood urea nitrogen, Hb = Hemoglobin, WBC = White blood cell count, Pro-BNP = N-terminal pro-brain natriuretic peptide, ESR = Erythrocyte sedimentation rate

Effects of ECV

Restoration of SR induced a significant reduction of heart rate (HR) and an improvement of shortening fraction and LV ejection fraction (LVEF). These changes were correlated with a postprocedural increase of LV end-diastolic volume (LVEDV) and a reduction of LV end-systolic diameter and volume [Table 2]. SR restoration determined a significant increase in endocardial longitudinal velocity in all segments, with the exception of the apical lateral wall. In 43 patients (89.6%), SR was associated with a marked improvement of peak LSt [Figure 2]. Post-ECV LSt was significantly correlated with pre-ECV LSt [Figure 3].

Table 2

Basal versus post-ECV clinical and echocardiographic evaluation

Variables	Evaluation
	Basal	Post-ECV	P-value
HR (b/min)	74±21	64±10	<0.001
SAP (mmHg)	143±23	132±20	<0.001
DAP (mmHg)	83±12	74±12	<0.001
Echocardiographic parameters
Left atrial diameter (mm)	48±7	49±8	0.419
Interventricular septum (mm)	10±1	11±2	0.253
LVEDD (mm)	53±9	52±9	0.154
Posterior wall (mm)	10±1	10±2	0.057
LVESD (mm)	37±10	35±10	0.021
LV shortening fraction (%)	31±10	34±10	0.017
LVEDV (mL/m2)	56±20	60±21	0.036
LVESV (mL/m2)	30±16	27±17	0.003
LV ejection fraction (%)	48±10	57±11	<0.001
Endocardial longitudinal velocity (cm/s)
Basal interventricular septum	3.3±0.8	4.3±1.3	<0.001
Mid interventricular septum	2.2±0.7	2.9±0.8	<0.001
Apical interventricular septum	1.1±0.4	1.3±0.5	0.040
Apical lateral wall	2.4±1.2	2.6±1.2	0.368
Mid lateral wall	3.0±1.1	3.5±1.4	0.007
Basal lateral wall	4.0±1.3	4.7±1.7	0.031

ECV = External cardioversion, HR = Heart rate, SAP/DAP = Systolic /diastolic arterial pressure, LV = Left ventricular, LVEDD/LVESD = LV end-diastolic/end-systolic diameter, LVEDV/LVESV = LV end-diastolic/end-systolic volume

Figure 2

Pre- and post-ECV values of peak longitudinal strain by segment explored. Restoration of sinus rhythm determines a significant improvement of peak longitudinal strain in any segment. Dotted lines indicate mean value at baseline and after the procedure

Figure 3

Correlation between pre- and post-ECV peak longitudinal strain (LSt) values (left panel) and between absolute changes of LSt (Δ) and pre-ECV LSt (right panel). In this last case, the dashed line indicates the “0” value

The variation of LSt (ΔLSt) induced by SR restoration correlated with the change of LVEF (β = −0.28 ± 0.06, R = 0.552, P < 0.001) and with the resumption of atrial mechanical activity observed in 44 patients (ΔLSt-A wave present: −5.4 ± 3.3% vs A wave absent: −0.8 ± 2.1%, P = 0.010). At univariate analysis [Tables 3 and 4, Figures 3 and 4], no association was found with pre-ECV LSt. Neither age nor gender showed a correlation with LSt changes. HF, aortic valve dysfunction, LV dilatation, dysthyroidism, and use of K+ sparing agents negatively correlated with LSt improvement. Basal HR and shortening fraction showed direct association with the absolute increase of strain. Interestingly, patients with a permanent pacemaker had the lower chance of improvement (ΔLSt-no pacemaker: −5.8 ± 3.2% vs single/dual chamber pacemaker with right ventricular stimulation [DDD/VVI]: −1.9 ± 2.9%, P = 0.008). IL-6 inversely correlated with the improvement of LV performance (ΔLSt −<0.16 pg/mL: −6.1 ± 3.5% vs 0.16-0.49 pg/mL: −5.0 ± 3.5% vs ≥0.50 pg/mL: −1.4 ± 2.2%, P = 0.019).

Table 3

Absolute change of peak longitudinal strain (ΔLSt) by categorical variables

Variables	Condition
	Present	Absent	P-value
Men	−6.3±3.4	−4.8±3.5	0.273
Smokers past-present	−5.6±3.3	−4.2±3.6	0.161
CAD	−5.2±3.5	−4.6±3.7	0.535
HF	−5.9±3.4	−3.3±3.1	0.011
Dyslipidemia	−5.4±3.4	−4.3±3.7	0.269
Hypertension	−5.2±4.4	−5.0±3.3	0.876
Valvular disease
Aortic	−5.6±3.5	−3.3±2.9	0.049
Mitral	−5.7±3.2	−4.4±3.7	0.122
CKD	−5.0±3.5	−5.3±3.7	0.908
COPD	−5.2±3.6	−4.2±3.4	0.526
CVD	−5.0±3.4	−4.8±4.6	0.897
Diabetes	−5.2±3.4	−3.6±4.4	0.279
Dysthyroidism	−5.6±3.5	−3.2±2.9	0.038
Therapy
ACE-I/ARB	−6.3±3.0	−4.7±3.6	0.200
Beta-blockers	−5.2±3.2	−4.9±3.7	0.739
Digoxin	−5.8±3.5	−3.9±3.2	0.060
Loop diuretics	−5.3±3.7	−4.4±3.1	0.391
K+ sparing agents	−5.4±3.3	−0.8±2.6	0.013
Statins	−5.0±3.4	−5.1±3.8	0.917
Oral anticoagulation	−7.3±2.1	−4.9±3.6	0.183
ASA	−4.9 ± 3.6	−6.1±2.9	0.406
Amiodarone	−5.8 ± 3.4	−4.2±3.4	0.102
Class I antiarrhythmic drugs	−5.0 ± 3.6	−5.1±3.4	0.950

CAD = Coronary heart disease, HF = Congestive heart failure, CKD = Chronic kidney disease, COPD = Chronic obstructive pulmonary disease, CVD = Cerebrovascular disease, ACE-I = Angiotensin converting enzyme inhibitors, ARB = Angiotensin receptor blockers, ASA = Acetylsalicylic acid

Table 4

Correlation between continuous variables and absolute change of peak longitudinal strain (ΔLSt; univariate linear regression analysis)

Variables	β + es	R	P-value
Weight (Δ·kg)	0.01±0.03	0.044	0.769
BSA (Δ·kg/m2)	−0.61±2.66	0.034	0.818
SAP (Δ·mmHg)	0.03±0.02	0.228	0.120
DAP (Δ·mmHg)	−0.01±0.04	0.050	0.734
BUN (Δ·mg/dL)	2.61±1.84	0.207	0.164
Creatinine (Δ·mg/dL)	1.99±1.28	0.226	0.127
K+(Δ·mEq/L)	−1.07±1.10	0.144	0.334
Hb (Δ·g/dL)	−0.20±0.33	0.088	0.552
WBC (Δ·n·10-3/mm3)	−0.10±0.35	0.046	0.773
Pro-BNP (Δ·mMol/L)	0.00±0.00	0.195	0.261
ESR (Δ·mm/h)	0.04±0.05	0.132	0.424
Echocardiographic parameters
Left atrial diameter (Δ·mm)	0.09±0.07	0.193	0.220
Interventricular septum (Δ·mm)	0.59±0.45	0.194	0.203
LVEDD (Δ·mm)	0.15±0.06	0.366	0.013
Posterior wall (Δ·mm)	0.25±0.38	0.100	0.514
LVESD (Δ·mm)	0.15±0.05	0.420	0.004
LV shortening fraction (Δ·%)	−0.12±0.05	0.337	0.024
LV ejection fraction (Δ·%)	−0.06±0.06	0.157	0.287

Δ = Absolute change of peak endocardial longitudinal strain after ECV by unitary variation of each independent variable, BSA = Body surface area, SAP/DAP = Systolic/diastolic arterial pressure, BUN = Blood urea nitrogen, Hb = Hemoglobin, WBC = White blood cell count, Pro-BNP = N-terminal pro-brain natriuretic peptide, ESR = Erythrocyte sedimentation rate, LV = Left ventricular, LVEDD/LVESD = LV end-diastolic/end-systolic diameter

Figure 4

Scatterplots of the univariate linear association of changes of peak longitudinal strain (dependent variable) with age, baseline values of heart rate (HR), and left ventricular (LV) end-diastolic and end-systolic volumes

At multivariate linear regression analysis, the resumption of atrial mechanical activity and baseline values of HR remained positively associated with LSt improvement. On the contrary, LVEDV and the presence of compensated dysthyroidism or a permanent pacemaker negatively correlated with ΔLSt [Table 5]. When IL-6 concentration (P = 0.020) was introduced into the model (R = 0.716, P < 0.001), HR (P = 0.952), LVEDV (P = 0.140), and dysthyroidism (P = 0.053) lost their statistical association with ΔLSt. To rule out the influence of artificial stimulation, we ran a further multivariate model which excluded patients with a permanent pacemaker: Baseline HR and LVEDV confirmed their correlation with changes of LSt (R = 0.551, P = 0.002).

Table 5

Clinical predictors of absolute change of peak longitudinal strain (ΔLSt)

Variables	β + es	95% CI	P-value
A wave (yes vs no)	−2.82±1.39	−(5.625-0.020)	0.048
Dysthyroidism (yes vs no)	+1.73±0.86	0.001-3.458	0.049
HR (Δ·b/min)	−0.04±0.02	−(0.079-0.001)	0.044
LVEDV (Δ·mL/m2)	+0.06±0.02	0.023-0.101	0.003
VVI/DDD (Δ vs No)	+2.95±1.04	0.847-5.063	0.007

Multivariate linear regression analysis; R = 0.729, P < 0.001. Excluded variables: Aortic valvular disease (P = 0.643), HF (P = 0.193), and K+ sparing agents (P = 0.304). ECV = External cardioversion, LV = Left ventricular, HR = Heart rate, A wave = presence of A wave after ECV, LVEDV = LV end-diastolic volume, VVI/DDD = Presence of a single/dual chamber pacemaker with right ventricular stimulation

DISCUSSION

Our study shows that effective ECV of AF results in a marked, immediate enhancement of LV performance, as demonstrated by the improvement of peak LSt. Lower baseline HR, lack of resumption of atrial mechanical activity, greater LV volumes, corrected alterations of thyroid function, presence of a pacemaker and an activated inflammatory status can identify a subset of patients at higher risk not to benefit from ECV.

LSt analysis increases sensitivity in detecting subclinical cardiac involvement in diabetic and hypertensive subjects.[24] Recent data suggest that LSt analysis has higher sensibility to detect changes in LV performance, compared to traditional echocardiography, also in AF patients.[25] This could justify the discrepancy in time course between the resumption of atrial (early) and ventricular (delayed) activity after successful ECV of chronic AF,[7] detected with a standard echocardiographic evaluation. Among 78 patients with persistent AF undergoing transcatheter ablation of the arrhythmia, strain measures improved while LVEF did not change.[25] Compared with a population in SR, patients with AF and normal LVEF had worse longitudinal and circumferential strain measures, which progressively normalized following successful ablation of the arrhythmia.[26] More recently, in a population of patients with acute myocardial infarction and preserved LVEF, LSt was an independent predictor of death or HF admissions over well-established risk factors such as age, diabetes, estimated glomerular filtration rate, hypertension, measures of infarct size and LV mass, and the degree of mitral regurgitation.[27]

Atrial dysfunction develops within few minutes from the appearance of AF.[28] A cardiac magnetic resonance imaging (MRI) study demonstrated that atrial and ventricular volumes significantly decreased after cardioversion.[29] In our study, after only 6 h from ECV, an effective atrial mechanical activity was observed in the vast majority of cases. This result was associated with a greater increase of LV performance, which could be explained by the dependence of LSt on loading condition.[24]

The improvement of LSt negatively correlated with LVEDV. This finding is in accordance with previous data which correlated the degree of ventricular dilatation with myocardial fibrosis in all types of cardiomyopathy.[3031] In a recent study, AF led to ventricular fibrosis only in dogs with normal atrioventricular (AV) conduction and a rapid ventricular response, compared to those treated with AV nodal ablation,[32] indicating that conversion to SR should be obtained as early as possible to avoid cardiac fibrosis and AF recurrence.[32]

In our multivariate analysis, IL-6 was so powerful to exclude HR and LVEDV from the model. The inverse relationship between ΔLSt and IL-6 may be mediated by fibrosis as well. Indeed, preclinical studies in rats have shown that intravenous infusion of IL-6 induced LV hypertrophy, increased wall stiffness, and collagen content.[33] In men, high concentrations of IL-6 directly correlated with left atrium dimension and length of AF.[34] More recently, in patients undergoing ECV of short-lasting AF, this cytokine was an independent predictor of arrhythmia recurrence.[35]

After ECV, the presence of a pharmacologically corrected thyroid dysfunction was linked to a reduced improvement of LSt. Indeed, a U-shaped relationship between myocardial strain and the level of thyroid hormones has been demonstrated in patients with differentiated thyroid carcinoma: During the transition from exogenous subclinical hyperthyroidism to overt hypothyroidism, circumferential and LSt progressively improved until hormonal profile was normal; but, soon after, decreased, reflecting progression to hypothyroidism.[36]

The presence of a permanent pacemaker was associated with a lower degree of LV improvement after ECV. This might be explained by the increased length of the pace-induced QRS, which can exacerbate further ventricular asynchrony. Indeed, in patients with sick sinus syndrome, compared to atrial pacing, ventricular pacing was associated with higher cardiovascular mortality, increased incidence of thromboembolic events and HF at follow-up.[37] In sinus node dysfunction and normal baseline QRS, right ventricular pacing increased the risk of hospitalization for HF and the incidence of AF even if AV synchrony was preserved.[38]

Study limitations

We enrolled patients with AF due to multiple causes which could have generated a nonhomogeneous response to effective ECV. However, our analysis did not identify disease-specific behaviors. Because all patients underwent correction of the arrhythmia by ECV, and electric shock usually produces a major degree of atrial stunning when compared to antiarrhythmic drugs,[39] this may limit the generalizability of our results. Finally, IL-6 concentration was measured only in a subset of patients.

CONCLUSION

Effective ECV of AF determines a prompt and significant improvement of LV performance. LSt assessment, a new tool for the analysis of cardiac function easily carried out at bedside, can effectively capture this change. In case of AF recurrence, LSt analysis may help directing the patient to a rhythm or a rate-control strategy based on the degree of improvement obtained after a previous effective ECV.