Clinical Investigations Left Atrial Function Assessed by Doppler Echocardiography Rather Than Left Atrial Volume Predicts Recurrence in Patients With Paroxysmal Atrial Fibrillation
Address for correspondence: Se-Joong Rim, MD, PhD and Jong-Youn Kim, MD, PhD Cardiology Division, Heart Center Gangnam Severance Hospital Yonsei University College of Medicine 146-92 Dogok-dong, Gangnam-gu Seoul, Korea 135-720
[email protected],
[email protected]
Ji Hyun Yoon, MD; Jeonggeun Moon, MD; Hye moon Chung, MD; Eui-Young Choi, MD, PhD; Jong-Youn Kim, MD, PhD; Pil-ki Min, MD, PhD; Young-won Yoon, MD, PhD; Byoung-kwon Lee, MD, PhD; Bum-kee Hong, MD, PhD; Hyuck-moon Kwon, MD, PhD; Se-Joong Rim, MD, PhD Cardiology Division, Department of Internal Medicine (Ji Hyun Yoon, Chung, Choi, Kim, Min, Young-won Yoon, Lee, Hong, Kwon, Rim), Yonsei University College of Medicine, Seoul, Republic of Korea; Cardiology Division, Department of Internal Medicine (Moon), Gacheon University Gil Hospital, Incheon, Republic of Korea
Background: Paroxysmal atrial fibrillation (PaAF) may present as a single self-terminating episode of atrial fibrillation (AF) or a more persistent form after sinus conversion. We investigated predictors of recurrence in patients with PaAF. Hypothesis: Left atrial function would be an useful parameter for predicting PaAF recurrence. Methods: The study population included 228 PaAF patients (131 males, age 64 ± 14 years) who underwent transthoracic echocardiography immediately after spontaneous sinus conversion at initial AF diagnosis. We followed the study patients for AF recurrence. Results: AF recurrence was demonstrated in 45 patients (20%, age 68 ± 13 years) after spontaneous sinus conversion. Patients with recurrence had larger left atrial volume index (32 ± 12 vs 25 ± 10 mL/m2 , P < 0.001), left ventricle mass index (107 ± 34 vs 93 ± 25 g/m2 , P = 0.012), and left ventricle filling pressure (E/e ) (14 ± 7 vs 12 ± 5, P = 0.012), whereas early diastolic mitral annular velocity (e ) (5 ± 2 vs 6 ± 3 cm/s, P = 0.021), late diastolic mitral annular velocity (A ) (7 ± 3 vs 9 ± 2 cm/s, P < 0.001), and peak systolic mitral annular velocity (7 ± 2 vs 8 ± 2 cm/s, P = 0.045) were significantly lower. In multivariate Cox regression analysis detecting independent predictors of PaAF recurrence, lower A (hazard ratio: 0.623, 95% confidence interval: 0.476-0.815, P = 0.001) was a significant predictor of AF recurrence. Conclusions: A , which indicates left atrial (LA) contractile function after sinus conversion, was the independent predictor of PaAF recurrence, whereas LA volume was not. LA function may be more important than LA volume in predicting recurrence particularly in patients with PaAF.
Introduction Paroxysmal atrial fibrillation (PaAF) has variable clinical progression after conversion to sinus rhythm. PaAF could present as a single self-terminating episode of atrial fibrillation (AF) in association with various systemic conditions.1 About 25% of PaAF will develop into a more persistent form at 5 years,2 with a recurrence rate of about 60% at 5 to 6 years.2 – 4 Most PaAF patients are maintained on antiarrhythmic oral anticoagulants and antiplatelet agents for a long time despite sinus rhythm maintenance. PaAF is associated with increased cardiovascular morbidity and mortality.1,5 Thus, determination of clinical and
The authors have no funding, financial relationships, or conflicts of interest to disclose. Received: November 23, 2012 Accepted with revision: January 30, 2013
echocardiographic parameters that might predict PaAF recurrence is essential for further improvement of treatment strategies and prognosis.6 However, clinical and echocardiographic predictors of PaAF recurrence still remain undetermined. Therefore, we sought to investigate predictors of PaAF recurrence and hypothesized that left atrial (LA) function would be a useful parameter for predicting PaAF recurrence.
Methods Study Population We retrospectively investigated 228 patients (131 males, age 64 ± 14 years) who were diagnosed with PaAF between January 2005 and October 2010 at our institution. Exclusion criteria were the presence of left ventricular ejection fraction Clin. Cardiol. 36, 4, 235–240 (2013) Published online in Wiley Online Library (wileyonlinelibrary.com) DOI:10.1002/clc.22105 © 2013 Wiley Periodicals, Inc.
235
(LVEF) of ≤40%, more than moderate valvular disease, congenital heart disease, hypertrophic cardiomyopathy, a permanent pacemaker, uncontrolled pulmonary disease, or thyroid dysfunction. Transthoracic echocardiography was performed after spontaneous sinus conversion in all patients at initial AF diagnosis. PaAF was defined as spontaneous termination of AF within 7 days.7 We followed the study patients for PaAF recurrence, which was defined as an episode of AF documented by electrocardiography (ECG) or 24-hour Holter monitoring after conversion to sinus rhythm. We divided the study population into 2 groups according to PaAF recurrence documented at least once on ECG or 24-hour Holter monitoring: group 1 (n = 45) with PaAF recurrence and group 2 (n = 183) without recurrence. Echocardiography All patients received a comprehensive echocardiographic examination (Sonos 5500, Philips or Vivid 7; GE Healthcare). Left ventricular end-diastolic diameter (LVEDD), left ventricular end-systolic diameter, and LA anteroposterior diameter were obtained from the 2-dimensional parasternal long-axis view. LA volume was determined by prolate ellipsoid formula and was indexed to the body surface area (LAVI). The ellipsoid model supposes that the LA can be measured as a prolate ellipse with a volume of 4π/3 (L/2) (D1/2) (D2/2), where L is the long axis (ellipsoid) and D1 and D2 are shortaxis dimensions.8 Left ventricular (LV) mass was calculated from LVEDD and wall thickness with the following formula: 0.8[1.04(SWTd + LVIDd + PWTd)3 -LVIDd3 ] + 0.6 (g), where LVIDd is the internal diameter of the LV, and PWTd and SWTd are posterior wall thickness and septal wall thickness at the end diastole, respectively.8 LVEF was calculated from maximal (LV end-systole just before mitral valve opening) and minimal (LV end-diastole at the start of QRS) LV volumes that were obtained by the modified Qui˜ nones method as described previously.9 Peak velocity of early diastolic filling (E), and late diastolic filling (A) and deceleration time of the E wave velocity were measured by the pulsed wave Doppler method at the level of the mitral valve leaflet tips on apical 4-chamber view. Peak early diastolic velocity (e ), late diastolic velocity (A ), and peak systolic velocity of mitral annulus (S ) were measured by tissue Doppler imaging (TDI) from the septal mitral annulus on apical 4-chamber view. E/e ratio was calculated to estimate LV filling pressure.10,11 The echocardiographic data were evaluated by 2 cardiologists who were blind to the patients clinical data. Statistical Analysis Continuous variables are presented as mean ± standard deviation and categorical variables as absolute numbers and percentages. Comparisons of baseline characteristics between the 2 groups were made by t test for continuous variables and χ2 test for categorical variables. A multivariate Cox regression analysis was performed to detect independent predictors of PaAF recurrence. Variables significant in the univariate analysis (P < 0.05) were included in the multiple Cox regression analysis. We used receiver
236
Clin. Cardiol. 36, 4, 235–240 (2013) J.H. Yoon et al: LA function and recurrence of PaAF Published online in Wiley Online Library (wileyonlinelibrary.com) DOI:10.1002/clc.22105 © 2013 Wiley Periodicals, Inc.
operating characteristics curves to examine cutoff values of predictors. We also used new statistical analysis methods to increase discriminative value of the new biomarker under assessment, A’ by calculation of integrated discrimination improvement (IDI), and net reclassification improvement (NRI) indices. NRI index is quantified as the net difference between the upward movement in categories for event subjects and the downward movement in those for nonevent subjects. A P value of
Male
131 (57)
Age, mean ± SD, y
Female
97 (43)
Underlying diseases, n (%)
Underlying diseases, n (%) Hypertension
110 (48)
Diabetes mellitus
35 (15)
Cerebrovascular accident
12 (30)
18(10.5)
0.004
Cerebrovascular accident
43 (17)
Alcohol, n (%)
7 (20)
36 (20)
0.672
9 (4)
Smoking, n (%)
8 (18)
40 (22)
0.684
History of cardiac surgery, n (%)
2 (4)
7 (4)
1.000
History of antiarrhythmic drugs, n (%)
11 (24)
38 (21)
0.108
Follow-up duration, mean ± SD, mo
13 ± 18
19 ± 31
0.183
History of cardiac operation Alcohol, n (%)
43 (17)
Smoking, n (%)
48 (19)
Antiarrhythmic drugs, n (%)
50 (22)
Follow-up duration, mo
18 ± 28 (0.3–119)
Echocardiographic parameters, mean ± SD (range) LVEDD, mm
46 ± 6 (25–71)
LVESD, mm
30 ± 5 (15–64)
EF, %
67 ± 8 (21–79)
LA diameter, mm 2
LAVI, mL/m , LVMI, g/m
2
37 ± 6 (24–53) 27 ± 11 (9.9–71) 95 ± 27 (40.8–202)
E wave, cm/s
68 ± 21 (24–146)
A wave, cm/s
72 ± 19 (24–127)
E/A
1 ± 0.4 (0.4–3.0)
6 ± 3 (2–16)
8 ± 2 (2–17)
e , cm/s A , cm/s
S , cm/s
7 ± 2 (3–14)
12 ± 6 (4.4–35.6)
E/e
e /A
0.8 ± 0.4 (0.1–2.3)
DT, ms
219 ± 53 (106–442)
RVSP, mm Hg
28 ± 10 (11–81)
Abbreviations: A, peak velocity of late diastolic filing; A , late diastolic mitral annular velocity; DT, deceleration time; E, peak velocity of early diastolic filling; e , early diastolic mitral annular velocity; EF, ejection fraction; LA, left atrial; LAVI, left atrial volume index; LVEDD, left ventricular end-diastolic diameter; LVESD, left ventricular end-systolic diameter; LVMI, left ventricular mass index; RVSP, right ventricular systolic pressure; S , peak systolic velocity of mitral annulus.
operating characteristics curve; area under curve = 0.706, P < 0.001). To find that the addition of atrial systolic function to LAVI improved the prediction of PaAF recurrence, we used a new statistical analysis methods (NRI and IDI indices). The addition of atrial systolic function to LAVI provided an IDI
Abbreviations: SD, standard deviation.
of 0.0558% (P = 0.001, standard error [SE]: 0.0173) and NRI of 0.552% (P = 0.001, SE: 0.1603).
Discussion The principal findings of this study are that PaAF recurrence was demonstrated in about 20% of patients, and the recurrence group had lower A velocity than the nonrecurrence group. Importance of LA Function in Patients With PaAF After Sinus Conversion Electrical and structural remodeling of the left atrium takes place with sustained AF. LA size also increases by scar formation, and a larger LA was thought to be a marker of impaired LA function. Larger LA size is certainly 1 of the independent risk factors of AF occurrence in patients with normal sinus rhythm,12 and a recent study showed LA function is suppressed in PaAF patients.13 However, the relationship between LA function and PaAF recurrence remains controversial. We hypothesized that preservation of LA function would reduce the chance of PaAF recurrence. In our study, patients with recurrent PaAF had a significantly lower A velocity compared to those without PaAF recurrence. We first demonstrated that LA function reflected by A velocity could be an useful predictor of PaAF recurrence. Diastolic Function Assessed by Doppler Echocardiography After Sinus Restoration Previous reports documented a pseudorestrictive pattern of diastolic function in PaAF patients after sinus conversion, which was considered unrelated to diastolic dysfunction but associated with decreased LA function expressed by a small Clin. Cardiol. 36, 4, 235–240 (2013) J.H. Yoon et al: LA function and recurrence of PaAF Published online in Wiley Online Library (wileyonlinelibrary.com) DOI:10.1002/clc.22105 © 2013 Wiley Periodicals, Inc.
237
Table 3. Intergroup Comparison of Echocardiographic Parameters Recurrence (n = 45), Mean ± SD
Nonrecurrence (n = 183), Mean ± SD
P Value
LVEDD, mm
47 ± 7
45 ± 5
0.145
LAVI, mL/m2
32 ± 12
25 ± 10
