The demographic data of the patients were examined in Table 1. No significant difference was found in demographic characteristics between the PAF and persistent AF groups. When the demographic characteristics of the patients and their association with post-procedural recurrence were examined (Table 1), significantly more AF/AT recurrences were observed in patients with persistent AF compared to those with PAF (p = 0.005). Additionally, the statistical significance of the association between higher CHA2DS2-VASc scores and AF/AT recurrence was demonstrated. The average duration from femoral sheath insertion to the end of the procedure was 60 ± 11.8 minutes.
Table 1
Demographic Data and Effects on Recurrence
Parameters | Study Population (n = 92) | Recurrence (n = 19) | No AF recurrence (n = 73) | p-value |
---|
Age | 59 ± 11.7 | 60,2 ± 8,4 | 56,4 ± 12,4 | 0,199 |
Male sex n(%) | 50 (54,3) | 8 (16) | 42 (84) | 0,145 |
HTN n(%) | 50 (54,3) | 13 (26) | 37 (74) | 0,280 |
DM n(%) | 23 (25) | 7 (30,4) | 16 (69,6) | 0,243 |
HLD n(%) | 12 (13,3) | 1 (8,3) | 11 (91,7) | 0,452 |
CAD history n(%) | 10 (10,9) | 1 (10) | 9 (90) | 0,685 |
CVA history n(%) | 1 (1,1) | 0 (0) | 1 (100) | 1 |
Pacemaker/ICD implantation n(%) | 3 (3,3) | 0 (0) | 3 (100) | 1 |
AAD use n(%) | 75 (81,5) | 17 (22,7) | 58 (77,3) | 0,651 |
Previous SVT Ablation n(%) | 3 (3,3) | 0 (0) | 3 (100) | 1 |
Paroxysmal AF n (%) | 73 (79,3) | 11 (15,1) | 62 (84,9) | 0,005 |
BMI (kg/m2) | 28,8 ± 4,6 | 30,2 ± 6,2 | 28,3 ± 4 | 0,109 |
CHA2DS2-VASc score | 1,7 ± 1,3 | 2,35 ± 1,3 | 1,56 ± 1,3 | 0,023 |
AF indicates atrial fibrillation; DM, diabetes mellitus; HLD, hyperlipidemia; HTN, hypertension; CAD, coronary artery disease; ICD, implantable cardioverter-defibrillator; CVA, cerebrovascular accident; BMI, Body Mass Index; AAD, Antiarrhythmic drug; SVT, supraventricular tachycardia
3.1. Biochemical Findings
The baseline levels of PRO-BNP, human ANP, PIIINP, MYBPHL, and hs-CRP were examined for each patient before the procedure (Table 2). Additionally, the MYBPHL value was assessed at 2 hours and 24 hours after the procedure.
Table 2
Baseline biomarker values in Paroxysmal and Persistent AF patients
Parameters | Study Population (n = 92) | Paroxysmal AF (n = 73) | Persistent AF (n = 19) | p-value |
---|
PRO-BNP (pg/mL) | 511,7 ± 666,5 | 365 ± 545,5 | 1074,2 ± 799,2 | < 0,005 |
Human ANP (pg/mL) | 274 ± 215,4 | 276,3 ± 233 | 265,2 ± 132,1 | 0,847 |
hs-CRP (mg/L) | 14,9 ± 18,6 | 15,6 ± 20,1 | 12,2 ± 10,8 | 0,492 |
PIIINP (pg/mL) | 1320,9 ± 604,4 | 1293,5 ± 532,7 | 1425,9 ± 835,7 | 0,411 |
MYBPHL 0 hours (ng/mL) | 15,9 ± 10,6 | 16,4 ± 11,3 | 13,8 ± 7,2 | 0,349 |
MYBPHL 2nd hour (ng/mL) | 44,3 ± 38,6 | 41,5 ± 38,7 | 54,8 ± 37,5 | 0,199 |
MYBPHL 24th hour (ng/mL) | 15 ± 11 | 14,3 ± 11,1 | 17,6 ± 10,7 | 0,274 |
Abbreviations: PIIINP: N-terminal pro-collagen type III propeptide, hs-CRP: high-sensitive C-reactive protein, BNP: brain natriuretic peptide, MYBPHL: myosin-binding protein H-like, AF: Atrial Fibrillation
When the biochemical values were examined in PAF and persistent AF subgroups (Table 2), the PRO-BNP value was statistically significantly higher in persistent AF patients compared to PAF patients (1074.2±799.2 vs 365±544.5; p<0.005). Although PIIINP, 2nd and 24th hour MYBPHL values were higher in the persistent AF group compared to the PAF group, they did not reach statistical significance. Consistent with the literature, MYBPHL values peaked at 2 hours post-procedure and returned to baseline values at 24 hours post-procedure.
The relationship between PRO-BNP, human ANP, PIIINP, hs-CRP, and 0-2-24-hour MYBPHL values and AF/AT recurrence was examined (Table 3). Although 0-2-24-hour MYBPHL values and hs-CRP values did not reach statistical significance, they were higher in patients with recurrence compared to those without recurrence. ROC analysis revealed that the predictive effect of MYBPHL values on recurrence was weak.
Further analysis revealed that the rise in MYBPHL biomarker from 0 hours to the peak at 2 hours post-procedure did not impact recurrence determination (34.7±4.2 vs 42.4±9.7; p=0.356). However, the PRO-BNP value was significantly higher in patients with recurrence compared to those without recurrence (851.9±723 vs 422.9±626.5; p=0.014).
Table 3: Effects of Biomarker Values on Recurrence
Parameters
|
Recurrence (n=19)
|
No AF recurrence (n=73)
|
p-value
|
PRO-BNP (pg/mL)
|
851,9±723
|
422,9±626,5
|
0,014
|
Human ANP (pg/mL)
|
265,5±132,3
|
276,3±234,1
|
0,848
|
hs-CRP (mg/L)
|
17,6±15,6
|
14,2±19,4
|
0,480
|
PIIINP (pg/mL)
|
1282,4±495,2
|
1331,6±634,5
|
0,755
|
MYBPHL 0 hours (ng/mL)
|
16,7±8,1
|
15,6±11,3
|
0,699
|
MYBPHL 2nd hour (ng/mL)
|
52,1±42,8
|
42,1±37,4
|
0,327
|
MYBPHL 24th hour (ng/mL)
|
15,3±7,2
|
14,9±11,9
|
0,904
|
Abbreviations: PIIINP: N-terminal pro-collagen type III propeptide, hs-CRP: high-sensitive C-reactive protein, BNP: brain natriuretic peptide, MYBPHL: myosin-binding protein H-like, AF: Atrial Fibrillation
3.2. Echocardiographic Findings
A comparative echocardiographic evaluation, including strain parameters, was conducted for each patient before the procedure and at the 3-month follow-up (Table 4). The measured LA diameter from the apical long-axis view was 41.1±5.8 mm before the procedure and 41.2±5.7 mm at the 3-month follow-up (p=0.191). There was no significant change observed in the left atrium volume index (LAVI) measured using the biplane method and indexed to body surface area. No significant changes were noted in the E and A wave velocities measured by the Doppler method in the 3-month follow-up echocardiogram compared to before the procedure. Similarly, there were no statistically significant changes in lateral and septal E/e' values, which are indicators of left ventricular diastolic dysfunction. Although the electromechanical delay measured from lateral and medial mitral annular tissue showed an increase at the 3-month follow-up compared to before the procedure, it did not reach statistical significance. In the analyses conducted using the strain method, the calculated LAsr (28.6±13.1 vs. 34.3±13.5; p<0.005), LAsct (-11.9±5.6 vs. -14.2±6.9; p=0.005), LAscd (-18±9 vs. -20.5±8.9; p=0.015), and LV endo average strain (-20.4±3.7 vs. -22±5.8; p=0.014) values showed a significant increase at the 3-month follow-up compared to before the procedure.
Table 4: Comparison of Pre-procedural and 3-Month Follow-up Echocardiographic Values for the Entire Patient Population
Parameters
|
Pre-procedure (n=92)
|
3-month follow up (n=92)
|
p value
|
LA diameter (mm)
|
41,1±5,8
|
41,2±5,7
|
0,191
|
LAVI (mL/m2)
|
30,6±11,4
|
29,8±11
|
0,199
|
LA-EMD medial (msn)
|
74,3±19,6
|
77,2±19,4
|
0,182
|
LA-EMD lateral (msn)
|
94,4±19,7
|
98,1±19
|
0,056
|
E/e’
|
9,4±2,7
|
9,5±3
|
0,700
|
Medial E/e’
|
10,1±3,1
|
10,2±3,6
|
0,715
|
Lateral E/e’
|
9,2±3,1
|
9,3±3,1
|
0,666
|
E wave velocity (m/sn)
|
83,6±20,2
|
83±20,3
|
0,767
|
A wave velocity (m/sn)
|
66±20,6
|
67,3±19
|
0,630
|
TRV (m/sn)
|
2,4±0,2
|
2,4±0,27
|
0,808
|
LAsr
|
28,6±13
|
34,3±13,5
|
<0,005
|
LAsct
|
-11,9±5,6
|
-14,2±6,9
|
0,005
|
LAscd
|
-18±9
|
-20,5±8,9
|
0,015
|
LV endo average strain
|
-20,4±3,7
|
-22±5,8
|
0,014
|
Abbreviations: LA: left atrium, LAVI: left atrial volume index, EMD: electromechanical delay, TRV: tricuspid regurgitation velocity, LAsr: left atrium reservoir strain, LAsct: left atrium contractile strain, LAscd: left atrium conduit strain, LV: left ventricle
In the paroxysmal AF group, echocardiographic changes were generally consistent with the overall patient population, although an increase was observed in LAscd (-19.6±8.9 vs. -21.3±8.9; p=0.129) and LV endo average strain (-21.5±2.9 vs. -22.5±6.3; p=0.193) values at 3 months, it did not reach statistical significance (Table 5).
Table 5: Comparison of echocardiographic values before the procedure and at the 3-month follow-up in paroxysmal and persistent AF patients
|
Paroxysmal AF population
|
Persistent AF population
|
Parameters
|
Pre-procedure (n=73)
|
3-month follow-up (n=73)
|
p-value
|
Pre-procedure (n=19)
|
3-month follow-up (n=19)
|
p-value
|
LA dimension (mm)
|
39,6±5,4
|
39,8±5,4
|
0,129
|
46,3±3,8
|
46,3±3,7
|
0,841
|
LAVI (mL/m2)
|
27,4±9,4
|
27,1±9,8
|
0,664
|
42,2±10,7
|
39,7±9,5
|
0,168
|
LA-EMD medial (msn)
|
72,5±19,6
|
75,4±18,9
|
0,182
|
88,1±14,1
|
90,7±18,5
|
0,734
|
LA-EMD lateral (msn)
|
91,5±18,3
|
95,3±17,6
|
0,052
|
116,3±16,9
|
119,2±16,1
|
0,643
|
E/e’
|
8,9±2,5
|
9,3±2,7
|
0,268
|
11,1±3,1
|
10,3±4
|
0,134
|
Medial E/e’
|
9,9±2,6
|
9,9±2,8
|
0,986
|
10,8±4,2
|
11,3±5,7
|
0,440
|
Lateral E/e’
|
9±3,1
|
9,1±3
|
0,597
|
9,9±3,3
|
9,9±3,7
|
0,939
|
E wave velocity (m/sn)
|
80,2±20
|
81,2±19,9
|
0,666
|
95,7±16,6
|
89,5±20,9
|
0,187
|
A wave velocity (m/sn)
|
67,5±20,3
|
66±18,3
|
0,540
|
54±20,2
|
77,9±22,6
|
0,061
|
TRV (m/sn)
|
2,38±0,24
|
2,39±0,27
|
0,605
|
2,6±0,25
|
2,5±0,24
|
0,229
|
LAsr
|
31,8±12,2
|
36,2±12,8
|
<0,005
|
16,5±8,5
|
26,7±13,6
|
0,015
|
LAsct
|
-12,9±5,5
|
-15,2±6,7
|
0,008
|
-7,9±4,4
|
-10,2±6,5
|
0,271
|
LAscd
|
-19,6±8,9
|
-21,3±8,9
|
0,129
|
-12,1±6,9
|
-17,6±8,3
|
0,035
|
LV endo average strain
|
-21,5±2,9
|
-22,5±6,3
|
0,193
|
-16,6±3,8
|
-20,3±3,6
|
<0,005
|
Abbreviations: LA: left atrium, LAVI: left atrial volume index, EMD: electromechanical delay, TRV: tricuspid regurgitation velocity, LAsr: left atrium reservoir strain, LAsct: left atrium contractile strain, LAscd: left atrium conduit strain, LV: left ventricle
In patients with persistent AF, unlike the general patient population, the increase in LAsct values at 3 months did not reach a statistically significant level (-7.9±4.4 vs. -10.2±6.5; p=0.271) (Table 5).
The changes in echocardiographic parameters of patients without recurrence were consistent with the general patient population (Table 6). Additionally, the 3-month follow-up LAVI value in this patient group was significantly lower compared to the pre-procedural value (28.9±10.5 vs 27.3±8.1; p=0.013).
Table 6: Comparison of pre-procedural and 3-month follow-up echocardiographic parameters in patients with and without recurrence after the procedure
|
No AF recurrence
|
AF recurrence
|
Parameters
|
Pre-procedure (n=73)
|
3-month follow-up (n=73)
|
p-value
|
Pre-procedure (n=19)
|
3-month follow-up (n=19)
|
p-value
|
LA dimension (mm)
|
40,4±5,9
|
40,4±5,8
|
1
|
43,5±4,6
|
44±4,8
|
0,002
|
LAVI (mL/m2)
|
28,9±10,5
|
27,3±8,1
|
0,013
|
36,3±12,9
|
38,6±14,6
|
0,006
|
LA-EMT medial (msn)
|
73,1±18,5
|
75,4±17,1
|
0,309
|
80,3±24,3
|
85,8±27,3
|
0,378
|
LA-EMT lateral (msn)
|
92,2±19,2
|
96,8±19,4
|
0,060
|
105,3±19,1
|
104,6±15,7
|
0,843
|
E/e’
|
9,1±2,5
|
9,3±2,7
|
0,605
|
10,4±3,3
|
10,3±3,9
|
0,787
|
Medial E/e’
|
10±2,7
|
9,9±2,9
|
0,795
|
10,6±4
|
11,3±5,4
|
0,155
|
Lateral E/e’
|
9±2,9
|
9,2±2,9
|
0,351
|
9,9±3,8
|
9,5±3,9
|
0,512
|
E wave velocity (m/sn)
|
81,1±18,7
|
80,5±19,6
|
0,807
|
92,1±23,4
|
91,5±20,6
|
0,859
|
A wave velocity (m/sn)
|
66,9±20,4
|
68±17,5
|
0,642
|
61,5±21,7
|
63,3±26
|
0,863
|
TRV (m/sn)
|
2,4±0,25
|
2,39±0,24
|
0,525
|
2,54±0,28
|
2,57±0,33
|
0,599
|
LAsr
|
30,5±12,4
|
37,3±12
|
<0,005
|
22,3±13,3
|
24,2±13,4
|
0,383
|
LAsct
|
-12,9±5,6
|
-15,5±6,3
|
0,006
|
-8,5±4,5
|
-9,7±7,3
|
0,423
|
LAscd
|
-18,7±8,7
|
-22±9,1
|
0,005
|
-16±10
|
-15,6±6,1
|
0,828
|
LV endo average strain
|
-20,8±3,5
|
-22,4±6,3
|
0,048
|
-19,2±4,1
|
-20,7±3,5
|
0,017
|
Abbreviations: LA: left atrium, LAVI: left atrial volume index, EMD: electromechanical delay, TRV: tricuspid regurgitation velocity, LAsr: left atrium reservoir strain, LAsct: left atrium contractile strain, LAscd: left atrium conduit strain, LV: left ventricle
In the group of patients with recurrence, a significant increase was observed in the 3-month LA diameter and LAVI values compared to pre-procedural values (43.5±4.6mm vs 44±4.8mm; p=0.002) (36.3±12.9 vs 38.6±14.6; p=0.006) (Table 6). Unlike the group of patients without recurrence, no significant change was observed in LA strain values in the recurrent group. However, a significant increase was observed in LV endo average strain value at 3 months (-19.2±4.1 vs -20.7±3.5; p=0.017).
The impact of demographic data, biochemical, and echocardiographic parameters on predicting recurrence in 92 atrial fibrillation (AF) patients was examined through logistic regression with univariate and multivariate analysis (Table 7). In the univariate analysis of patients, AF type (p=0.004), CHA2DS2-VASc score (p=0.027), PRO-BNP value (p=0.022), LA diameter (p=0.035), LAsr (p=0.018), LAsct (p=0.005), LA-EMT lateral duration (p=0.018), and E-wave velocity (p=0.033) were found to be significantly associated with recurrence. In the multivariate analysis of patients, PRO-BNP value (p=0.012), LAsr (p=0.047), and LAsct (p=0.024) were independently associated with recurrence.
Table 7: Univariate and Multivariate Analysis with Logistic Regression
Univariate and multivariate analyses for recurrence
|
|
Univariate
|
Multivariate
|
|
CI
|
OR(ExpB)
|
P-value
|
CI
|
OR(ExpB)
|
P-value
|
Demographic Data
|
AF type (Paroxysmal:1, persistant:2)
|
1,679-15,32
|
5,073
|
0,004
|
0,072-52
|
1,940
|
0,693
|
CHA2DS2-VASc
|
1,048-2,208
|
1,521
|
0,027
|
0,928-3,231
|
1,731
|
0,085
|
Biochemical Finding
|
PRO-BNP
|
1,000-1,002
|
1,001
|
0,022
|
1,000-1,003
|
1,002
|
0,012
|
Echocardiographic Results
|
LA dimension
|
1,007-1,211
|
1,104
|
0,035
|
0,825-1,221
|
1,004
|
0,970
|
LAsr
|
0,901-0,990
|
0,945
|
0,018
|
1,004-1,246
|
1,113
|
0,047
|
LAsct
|
1,056-1,369
|
1,202
|
0,005
|
1,040-1,750
|
1,349
|
0,024
|
LV endo average strain
|
0,987-1,281
|
1,124
|
0,078
|
0,794-1,479
|
1,084
|
0,611
|
LA-EMD lateral
|
1,006-1,069
|
1,037
|
0,018
|
0,963-1,066
|
1,013
|
0,622
|
E wave velocity
|
1,002-1,058
|
1,030
|
0,033
|
0,960-1,056
|
1,007
|
0,782
|
Abbreviations: LA: left atrium, LAVI: left atrial volume index, EMD: electromechanical delay, TRV: tricuspid regurgitation velocity, LAsr: left atrium reservoir strain, LAsct: left atrium contractile strain, LAscd: left atrium conduit strain, LV: left ventricle, AF: atrial fibrillation