Circulation Journal
Online ISSN : 1347-4820
Print ISSN : 1346-9843
ISSN-L : 1346-9843
Arrhythmia/Electrophysiology
Elevated Red Blood Cell Distribution Width Predicts Recurrence After Catheter Ablation for Atrial Fibrillation in Patients With Heart Failure　– Comparison With Non-Heart Failure Patients –
Satoshi Yanagisawa, Yasuya Inden, Hiroyuki Kato, Aya Miyoshi, Yoshiaki Mizutani, Tadahiro Ito, Yosuke Kamikubo, Yasunori Kanzaki, Makoto Hirai, Toyoaki Murohara
Author information
Keywords: Atrial fibrillation, Catheter ablation, Heart failure, Red blood cell distribution width
JOURNAL FREE ACCESS FULL-TEXT HTML
Supplementary material

2016 Volume 80 Issue 3 Pages 627-638

Browse “Advance online publication” version
Details
Download PDF (1071K)
Download citation RIS

(compatible with EndNote, Reference Manager, ProCite, RefWorks)

BIB TEX

(compatible with BibDesk, LaTeX)

Text
How to download citation
Contact us
Abstract

Background: Elevated red blood cell distribution width (RDW) predicts poor prognosis in patients with cardiovascular diseases. However, little is known about the association between RDW and outcomes after catheter ablation of atrial fibrillation (AF).

Methods and Results: A total of 757 patients who underwent radiofrequency catheter ablation of AF were divided into heart failure (HF, n=79) and non-HF (n=678) groups; RDW was assessed as a predictor after catheter ablation in each. During a 22.3-month follow-up period, the baseline RDW in the HF group was greater in the recurrence group than in the non-recurrence group (14.5±2.0% vs. 13.5±0.9%, P=0.013). In contrast, no significant difference in RDW at baseline was found in the non-HF group between the recurrence and non-recurrence groups (13.3±0.8% vs. 13.2±0.8%, P=0.332, respectively). Multivariate analysis demonstrated that RDW (hazard ratio 1.20, 95% confidence interval 1.01–1.40, P=0.034) was an independent predictor of AF recurrence in the HF group. The cut-off values of RDW for the recurrence of AF and major adverse events in the HF group were 13.9% and 14.8%, respectively.

Conclusions: High RDW is an independent predictor for the recurrence of AF and major adverse events in patients with HF after catheter ablation. RDW is a potential noninvasive marker in AF patients complicated with HF. (Circ J 2016; 80: 627–638)

Red blood cell distribution width (RDW) is a measurement parameter of the variability of circulating red blood cell size and is easily available as part of a standard complete blood cell count. Traditionally, RDW is used as a method of differential diagnosis of anemia etiology.1 High RDW reflects an elevated reticulocyte level because of the possibility of iron deficiency, hemolytic anemia, or folate and vitamin B deficiency. Recently, several studies have reported that elevated RDW is a prognostic factor for mortality and major adverse events in patients with heart failure (HF) and cardiovascular disease.2–6

As for atrial fibrillation (AF), a common cardiovascular disease, previous reports similarly demonstrated significant relationships between AF incidence and elevated RDW for prognostic assessment in the general population and postoperative patients.7,8 However, few data exist for the assessment of the relationship between RDW and outcomes in patients undergoing catheter ablation of AF. Recent studies with relatively small samples reported that an elevated RDW predicted a late recurrence of AF in patients undergoing cryoballoon-based ablation.9,10 No data exist from large-sample studies to evaluate the association between RDW and outcomes after catheter ablation, and the details of the possible mechanism underlying the abovementioned association remain unknown.

The present study sought to examine the association between RDW and outcomes after radiofrequency catheter ablation (RFCA) for AF in a large sample. Because RDW is mainly reported as a prognostic parameter of mortality and morbidity in patients with HF, we focused on the outcomes in patients with HF who underwent RFCA of AF, compared with non-HF patients, and individually assessed RDW as a prognostic factor after CA in this study.

Methods

Study Population

The study population was retrospectively recruited from a CA database at Nagoya University Hospital. The database was approved by the institutional ethical committee. Patients who underwent RFCA of AF for the first time between January 2009 and December 2014 were included in this study. All the patients were referred for CA because they were refractory to antiarrhythmic and rate-control drugs or could not be administered these drugs because of severe HF, side effects, or comorbidities. The indications for CA of AF complied with the latest guideline.11 Exclusion criteria were: (1) insufficient examination results at baseline or loss to follow-up within 3 months after CA; (2) emergency CA of AF for hemodynamic instability, and decompensated HF; (3) severe renal dysfunction with estimated glomerular filtration rate (eGFR) ≤30 ml/min/1.73 m2 at baseline; (4) history of CA or MAZE procedure; (5) development of a major complication resulting in discontinuation of the ablation procedure; and (6) abnormal thyroid function, severe anemia (hemoglobin <10.0 g/dl for men; <9.0 g/dl for women), hematological disease, liver cirrhosis, therapy affecting bone-marrow suppression, folate and vitamin B12 deficiency, history of recent blood transfusion, infection, hemorrhage events, and surgical operation. Patients with HF were defined as having a history of HF hospitalization or left ventricular ejection fraction (LVEF) ≤40% on baseline echocardiography.

Examination Course

Patients who were scheduled for CA were admitted the day before the procedure. Informed consent was given by all of the patients according to hospital guidelines. At admission, baseline blood testing, echocardiography, electrocardiography, and Holter examination were performed. Antiarrhythmic agents were stopped 5 half-lives before ablation, except for amiodarone and bepridil, which were stopped >1 week before the procedure. Transesophageal echocardiography was performed in all patients to confirm the absence of atrial thrombus. Anticoagulant drugs, including novel anticoagulant agents, were continued during the procedure, as previously reported.12

Ablation Procedure

In the ablation procedure, vascular access was obtained via the right and left femoral and left subclavian veins. We also accessed the right femoral artery to perform coronary angiography and monitor blood pressure. After transseptal puncture using intracardiac echocardiography, 3 sheaths (2 8F sheaths and 1 8.5Fr steerable sheath) were introduced into the left atrium. Then, using a circular mapping catheter (LassoTM, Biosense Webster Inc, Diamond Bar, CA, USA) placed on the ostium of each pulmonary vein (PV) atrium, encircling PV isolation was performed with a 3.5-mm tip, open-irrigated ablation catheter (Biosense Webster Inc) to achieve electric isolation of the PV potential. All ablation procedures were performed with a 3D electroanatomical mapping system (CARTOTM, Biosense Webster Inc). The RF energy output was titrated to 25–35 W at a flow rate of 17–30 ml/min, with a maximum temperature of 42℃. For the most part, paroxysmal AF and early persistent AF required PV isolation alone, but in patients with prolonged persistent AF, atrial tachycardia, or evidence of non-PV foci, additional linear ablation and complex fractionated electrogram ablation were applied. If the patient did not convert to sinus rhythm at the end of the ablation procedure, internal cardioversion was performed. During the procedure, bolus and additional heparin were administered to maintain an activated clotting time of 300–350s.

Follow-up

Patients remained hospitalized under continuous rhythm monitoring for 3 days after the procedure. After discharge, patients were followed through the outpatient clinic at 1, 3, 6 months, and every 6 months after ablation. At one month after ablation, 24-hour Holter monitoring was performed in all patients. At the time of each follow-up visit, patients underwent 12-lead ECG, and were asked about any symptoms related to the presence of arrhythmia. If patients were suspected of having had an emerging arrhythmia, but had no evidence of the arrhythmia at the time of examination, additional Holter monitoring and short-duration follow-up were performed. Device interrogation was also used to detect any recurrence of AF or atrial tachycardia in patients with a history of device implantation. AF or atrial tachycardia occurring within 3 months (blanking period) after ablation was not considered to be a recurrence; however, repeat ablation during the blanking period was defined as AF recurrence. If the patient had an AF episode during the blanking period, antiarrhythmic drugs that had been discontinued before the procedure were re-administered. If no AF episode occurred beyond the blanking period after administration of antiarrhythmic drug therapy, successful ablation without recurrence was documented. Discontinuation of antiarrhythmic agents was decided on the basis of freedom from recurrence of any atrial arrhythmia for more than 3–6 months’ follow-up after ablation. The primary follow-up endpoint of this study was the recurrence of any AF or atrial tachycardia of more than 30 s duration. Moreover, we defined major adverse events as all-cause death, HF hospitalization, and cerebral infarction after discharge of the catheter ablation procedure.

Laboratory Assessment and Other Testing

Blood samples were collected from a peripheral vein while the patient was supine after a rest period. The complete blood cell count and RDW were assessed using an XE-2100 automated hematology analyzer (Sysmex Inc, Kobe, Japan). The reference range for normal RDW values was 11.0–14.0%. The plasma B-type natriuretic peptide (BNP) concentration was measured with a specific immunoassay for human BNP (ARCHITECT BNP-JP kit, Abbott Japan Inc, Chiba, Japan). High-sensitivity C-reactive protein (hs-CRP) level was measured using human anti-CRP (CRP II Latex X2, Denka Seiken, Tokyo, Japan). The blood tests, including RDW, were performed on admission the day prior to ablation.

The CHADS2 and CHA2DS2-VASc scores were evaluated as previously described.13,14 The eGFR was calculated on the basis of the Japanese coefficient-modified Modification of Diet in Renal disease study equation.15 For the echocardiography parameters, LVEF was calculated using Simpson’s method. The left atrial diameter (LAD), left ventricular end-diastolic and end-systolic diameters were assessed using M-mode methods. This retrospective study was performed in accordance with the Declaration of Helsinki. The patients’ baseline characteristics, comorbidities, and therapeutic details were obtained from hospital medical records.

Statistical Analysis

Continuous variables are expressed as the mean±standard deviation or median (1st and 3rd quartiles), and categorical variables are expressed as number and percentage. Comparison of the differences in the baseline characteristics were performed using Student’s t test for parametric data and Mann-Whitney U-tests for non-parametric data. Categorical variables were compared using the chi-square test or Fisher’s exact test. The Kaplan-Meier method was used to estimate event-free survival, and the differences between the curves were compared using the log-rank test. The prognostic value of each factor was first evaluated by univariate Cox proportional hazard regression analysis. The factors that had P-values <0.05 in the univariate analysis were entered into a multivariate Cox proportional hazards model to identify the independent predictors. Based on the obtained significant predictors, a receiver-operating characteristic (ROC) curve was plotted, and the cut-off point for the ROC curve factor was determined. Bland-Altman difference plots with 95% confidence limits were constructed to evaluate the degree of agreement between the 2 measurements.16 The 95% limits of agreement were calculated as the mean difference±2 standard deviations. P<0.05 was considered statistically significant.

Results

A total of 757 patients were included in the present study. Of them, 79 were in the HF group, and 678 were in the non-HF group. Baseline characteristics and examination results between the HF and non-HF groups are shown in Table 1. The HF patients had a lower prevalence of paroxysmal AF and higher prevalence of persistent AF and a history of device implantation. In the laboratory data, white blood cell count, hs-CRP levels, RDW, creatinine levels, eGFR, and plasma BNP levels were higher in the HF group than in the non-HF group. Significant differences were observed between the 2 groups in echocardiographic parameters, CHADS2 and CHA2DS2-VASc scores, prevalence of medication therapies, and ablation procedures.

Table 1. Comparison of the Demographic and Baseline Characteristics of the Non-HF and HF Groups
  Non-HF group
(n=678)		 HF group
(n=79)		 P value
Age, years 61.4±11.6 63.6±11.2 0.108
Male sex 512 (76%) 58 (73%) 0.682
BMI, kg/m2 24.3±6.9 23.6±3.7 0.386
Type of AF
　Paroxysmal 476 (70%) 30 (38%) <0.001　
　Persistent 142 (21%) 37 (47%) <0.001　
　Long-standing persistent 60 (9%) 12 (15%) 0.069
Duration of AF, years 2.0 (0.4–5.1) 0.8 (0.4–5.3) 0.198
Comorbidities
　Hypertension 308 (45%) 30 (38%) 0.207
　Diabetes mellitus 86 (13%) 16 (20%) 0.062
　Coronary artery disease 48 (7%) 8 (10%) 0.327
　Stroke or TIA 56 (8%) 5 (6%) 0.551
Previous device implantation 16 (2%) 13 (17%) <0.001　
Laboratory data
　Hemoglobin, g/dl 14.0±1.5 13.7±2.1 0.073
　WBC count (×103 μl) 5.4±1.5 5.9±1.6 0.007
　hs-CRP, mg/L 0.50 (0.30–1.10) 0.90 (0.40–2.02) <0.001　
　RDW, % 13.2±0.8 14.0±1.6 <0.001　
　MCV, fl 92.1±4.3 91.8±5.6 0.476
　Creatinine level, mg/dl 0.8±0.2 1.0±0.2 <0.001　
　　eGFR, ml/min/1.73 m2 74.5±20.6 61.3±17.1 <0.001　
　BNP level, pg/dl 40.7 (21.1–88.3) 148.8 (70.0–289.0) <0.001　
　Total cholesterol, mg/dl 192.5±34.5 189.9±36.2 0.526
　Albumin, g/dl 4.2±0.3 4.1±0.4 0.096
Echocardiographic data
　LAD, mm 38.5±6.3 42.9±7.7 <0.001　
　LVEDD, mm 48.9±5.0 53.4±9.1 <0.001　
　LVEDS, mm 32.3±4.6 40.7±10.8 <0.001　
　LVEF, % 62.1±6.7 46.4±15.1 <0.001　
CHADS2 score 0.9±1.0 1.9±1.0 <0.001　
CHA2DS2-VASc score 1.6±1.5 2.8±1.5 <0.001　
NYHA functional class NA 2.1±0.7 NA
Medical therapy
　ACEI or ARB 236 (35%) 46 (58%) <0.001　
　β-blockers 210 (31%) 66 (84%) <0.001　
　Spironolactone 13 (2%) 49 (62%) <0.001　
　Diuretic 23 (3%) 61 (77%) <0.001　
　Digoxin 42 (6%) 17 (22%) <0.001　
Ablation procedure
　Pulmonary vein isolation 678 (100%) 79 (100%) NA
　Cavotricuspid isthmus 568 (84%) 70 (89%) 0.264
　LA linear ablation 211 (31%) 43 (54%) <0.001　
　CFAE 113 (17%) 28 (35%) <0.001　
　Superior vena cava isolation 54 (8%) 9 (11%) 0.297

The data are presented as number (%), and mean±standard deviation or median (interquartile). Linear ablation includes roof, bottom, and mitral isthmus lines. ACEI, angiotensin-converting enzyme inhibitor; AF, atrial fibrillation; ARB, angiotensin-receptor blocker; BMI, body mass index; BNP, B-type natriuretic peptide; CFAE, complex fractionated electrogram; eGFR, estimated glomerular filtration rate; HF, heart failure; hs-CRP, high-sensitivity C-reactive protein; LA, left atrial; LAD, LA diameter; LVEDD, left ventricular end-diastolic diameter; LVEF, left ventricular ejection fraction; LVEDS, left ventricular end-systolic diameter; MCV, mean corpuscular volume; NYHA, New York Heart Association; RDW, red blood cell distribution width; TIA, transient ischemic attack; WBC, white blood cell.

During a follow-up period (mean=22.3±16.7 months) after catheter ablation, recurrence of AF was observed in 311 patients (41%) in the total population. There was a significant difference in recurrence of AF between the HF and non-HF groups (42 patients [53%] vs. 269 patients [40%], P=0.021). Repeat CA was performed in 213 patients (HF group; 24 patients [30%] vs. non-HF group; 189 patients [28%], P=0.640) during the study period. During a repeat ablation procedure, a PV reconnection was found in 18 (75%) and 161 (85%) patients (P=0.234), and the mean number of PV reconnections was 2.1±1.0 and 2.4±1.0 (P=0.139) in the HF and non-HF groups, respectively. Class I and Class III antiarrhythmic drugs were administered to 160 (133 during a blanking period) and 72 patients (66 during a blanking period) in the total population.

Within the HF group, a comparison of the baseline characteristics and examination data of the recurrence and non-recurrence groups is shown in Table 2. Female sex and reduced prevalence of persistent AF were more common in the recurrence group than in the non-recurrence group. The baseline RDW was greater in the recurrence group than in the non-recurrence group (14.5±2.0% vs. 13.5±0.9%, P=0.013). Moreover, a significant difference in LAD was observed. In contrast, for the non-HF group, prevalence of long-standing persistent AF and BNP levels were higher in the recurrence group than in the non-recurrence group. On echocardiography, LAD, and the LV end-diastolic and -systolic diameters were greater in the recurrence group. No significant difference in RDW at baseline was observed between the recurrence and non-recurrence groups (13.3±0.8% vs. 13.2±0.8%, P=0.332).

Table 2. Comparison of the Demographic and Baseline Characteristics of the Non-Recurrence and Recurrence Groups of Non-HF and HF Patients
  Non-HF group (n=678) P value HF group (n=79) P value
Non-recurrence
(n=409)		 Recurrence
(n=269)		 Non-recurrence
(n=37)		 Recurrence
(n=42)
Age, years 61.6±12.0 61.1±11.1 0.523 63.0±10.6 64.2±11.8 0.647
Male sex 305 (75%) 202 (77%) 0.481 32 (87%) 26 (62%) 0.014
BMI, kg/m2 24.3±8.3 24.3±3.7 0.999 24.0±3.2 23.3±4.0 0.340
Type of AF
　Paroxysmal 295 (72%) 181 (67%) 0.178 10 (27%) 20 (48%) 0.060
　Persistent 87 (21%) 55 (20%) 0.796 23 (62%) 14 (33%) 0.010
　Long-standing persistent 27 (7%) 33 (12%) 0.011 4 (11%) 8 (19%) 0.309
Duration of AF, years 2.0
(0.4–5.0)		 2.0
(0.4–6.0)		 0.110 0.8
(0.4–2.6)		 0.9
(0.4–7.3)		 0.327
Comorbidities
　Hypertension 183 (45%) 125 (47%) 0.660 18 (47%) 12 (29%) 0.067
　Diabetes mellitus 54 (13%) 32 (12%) 0.617 6 (16%) 10 (24%) 0.402
　Coronary artery disease 30 (7%) 18 (7%) 0.749 2 (5%) 6 (14%) 0.271
　Stroke or TIA 27 (7%) 29 (11%) 0.053 1 (3%) 4 (10%) 0.364
Previous device implantation 10 (2%) 6 (2%) 0.857 3 (8%) 10 (24%) 0.060
Laboratory data
　Hemoglobin, g/dl 14.0±1.5 14.0±1.5 0.741 14.1±1.8 13.3±2.3 0.092
　WBC count (×103 μl) 5.3±1.6 5.5±1.4 0.273 6.1±1.6 5.7±1.5 0.311
　hs-CRP, mg/L 0.50
(0.20–1.10)		 0.50
(0.30–1.10)		 0.327 1.00
(0.40–2.20)		 0.80
(0.40–2.00)		 0.598
　RDW, % 13.2±0.8 13.3±0.8 0.332 13.5±0.9 14.5±2.0 0.013
　MCV, fl 92.0±4.2 92.3±4.4 0.407 92.3±4.6 91.3±6.4 0.432
　Creatinine level, mg/dl 0.8±0.2 0.8±0.2 0.338 1.0±0.2 1.0±0.2 0.880
　　eGFR, ml/min/1.73 m2 74.4±16.0 74.5±26.1 0.950 62.8±16.2 60.0±18.0 0.481
　BNP level, pg/dl 38.0
(19.4–83.8)		 47.8
(23.7–93.2)		 0.039 142.0
(59.2–264.2)		 165.2
(87.4–329.6)		 0.369
　Total cholesterol, mg/dl 193.3±34.7 191.4±34.1 0.494 190.4±32.2 189.5±39.8 0.905
　Albumin, g/dl 4.1±0.4 4.2±0.3 0.921 4.2±0.4 4.0±0.5 0.277
Echocardiographic data
　LAD, mm 38.1±6.2 39.2±6.4 0.018 40.9±7.3 44.7±7.8 0.032
　LVEDD, mm 48.5±4.9 49.7±4.9 0.003 53.9±6.4 53.1±11.1 0.704
　LVEDS, mm 31.8±4.5 32.9±4.6 0.001 40.7±8.0 40.7±12.9 0.970
　LVEF, % 62.4±6.8 61.6±6.5 0.164 45.8±15.1 46.9±15.2 0.757
CHADS2 score 0.9±1.0 0.9±1.1 0.501 1.8±0.8 1.9±1.3 0.707
CHA2DS2-VASc score 1.7±1.5 1.6±1.5 0.675 2.5±1.2 3.0±1.7 0.133
NYHA functional class NA NA NA 2.2±0.7 1.9±0.7 0.097
Medical therapy
　ACEI or ARB 142 (35%) 92 (35%) 0.952 20 (54%) 26 (62%) 0.480
　β-blockers 120 (29%) 90 (34%) 0.257 29 (78%) 37 (88%) 0.245
　Spironolactone 8 (2%) 5 (2%) 0.928 25 (68%) 24 (57%) 0.341
　Diuretic 15 (4%) 8 (3%) 0.626 28 (76%) 33 (79%) 0.759
　Digoxin 23 (6%) 19 (7%) 0.447 10 (27%) 7 (17%) 0.264
Ablation procedure
　Pulmonary vein isolation 409 (100%) 269 (100%) NA 37 (100%) 42 (100%) NA
　Cavotricuspid isthmus 343 (84%) 225 (84%) 0.939 31 (84%) 39 (93%) 0.292
　LA linear ablation 130 (32%) 81 (30%) 0.672 23 (62%) 20 (48%) 0.195
　CFAE 66 (16%) 47 (18%) 0.648 13 (35%) 15 (36%) 0.957
　Superior vena cava isolation 25 (6%) 29 (11%) 0.028 7 (19%) 2 (5%) 0.075
Antiarrhythmic drug use at follow-up
　Class I 53 (13%) 97 (36%) <0.001 1 (3%) 9 (21%) 0.012
　Class III 14 (3%) 41 (15%) <0.001 7 (19%) 10 (24%) 0.598
　None 342 (84%) 131 (49%) <0.001 29 (78%) 23 (55%) 0.027

Data are presented as number (%), and mean±standard deviation or median (interquartile). Abbreviations as in Table 1.

Univariate Cox proportional hazard regression analysis demonstrated that duration of AF, Class III antiarrhythmic drug administration at follow-up, stroke, BNP level, LAD, and LV end-diastolic and -systolic diameters were significantly associated with recurrence in the non-HF group. In contrast, female sex, RDW, and BNP levels were significant factors in recurrence in the HF group. Multivariate analyses showed that Class III antiarrhythmic drug use (hazard ratio [HR] 2.27, 95% confidence interval [CI] 1.59–3.24, P=0.001), stroke (HR 1.62, 95% CI 1.10–2.39, P=0.014), and BNP level (HR 1.00, 95% CI 1.00–1.01, P=0.010) were independent predictors of AF recurrence in the non-HF group, while female sex (HR 2.43, 95% CI 1.22–4.88, P=0.012) and RDW (HR 1.20, 95% CI 1.01–1.40, P=0.034) were independent predictors for AF recurrence in the HF group (Table 3).

Table 3. Predictors of Baseline Characteristics for Recurrence of AF in Univariate and Multivariate Regression Analyses in the Non-HF and HF Groups
  Non-HF group HF group
Univariate analysis Multivariate analysis Univariate analysis Multivariate analysis
HR (95% CI) P value HR (95% CI) P value HR (95% CI) P value HR (95% CI) P value
Age, year 0.99
(0.99–1.01)		 0.470     1.01
(0.98–1.03)		 0.662    
Female 0.87
(0.66–1.16)		 0.341     2.86
(1.51–5.45)		 0.001* 2.43
(1.22–4.88)		 0.012*
BMI, kg/m2 1.00
(0.99–1.02)		 0.814     0.93
(0.84–1.01)		 0.098    
Non-PAF 1.21
(0.94–1.56)		 0.149     0.62
(0.34–1.15)		 0.131    
Duration of AF, years 1.02
(1.00–1.04)		 0.046* 1.02
(0.99–1.04)		 0.073 1.03
(0.97–1.09)		 0.302    
Class III antiarrhythmic drug
use at follow-up		 2.67
(1.91–3.72)		 0.001* 2.27
(1.59–3.24)		 0.001* 1.40
(0.68–2.85)		 0.363    
Hypertension 1.08
(0.85–1.38)		 0.519     0.62
(0.32–1.20)		 0.156    
Diabetes mellitus 0.98
(0.68–1.42)		 0.924     1.51
(0.73–3.11)		 0.264    
Coronary artery disease 0.90
(0.56–1.45)		 0.666     1.12
(0.46–2.66)		 0.816    
Stroke or TIA 1.57
(1.07–2.31)		 0.022* 1.62
(1.10–2.39)		 0.014* 2.36
(0.83–6.66)		 0.106    
Hemoglobin, g/L 0.98
(0.90–1.05)		 0.515     0.88
(0.75–1.03)		 0.105    
WBC count, 103 μl 1.03
(0.96–1.11)		 0.444     0.91
(0.74–1.12)		 0.373    
hs-CRP, mg/L 0.92
(0.66–1.29)		 0.630     1.22
(0.84–1.76)		 0.290    
RDW, % 1.07
(0.93–1.23)		 0.378     1.30
(1.12–1.52)		 0.001* 1.20
(1.01–1.40)		 0.034*
MCV, fl 1.01
(0.98–1.04)		 0.465     0.96
(0.89–1.02)		 0.196    
Creatinine level, mg/dl 1.52
(0.79–2.91)		 0.209     1.31
(0.31–5.58)		 0.713    
eGFR, ml/min/1.73 m2 0.99
(0.99–1.01)		 0.870     0.99
(0.97–1.01)		 0.270    
BNP level, pg/dl 1.00
(1.00–1.01)		 0.001* 1.00
(1.00–1.01)		 0.010* 1.01
(1.00–1.01)		 0.037* 1.00
(1.00–1.00)		 0.442
Total cholesterol, mg/dl 1.00
(0.99–1.00)		 0.063     1.00
(0.99–1.01)		 0.925    
Albumin, g/dl 1.08
(0.77–1.53)		 0.662     0.73
(0.37–1.40)		 0.340    
LAD, mm 1.02
(1.00–1.04)		 0.020* 0.99
(0.97–1.02)		 0.591 1.03
(0.99–1.07)		 0.136    
LVEDD, mm 1.03
(1.01–1.06)		 0.009* 1.01
(0.97–1.06)		 0.547 0.99
(0.95–1.03)		 0.683    
LVEDS, mm 1.04
(1.02–1.07)		 0.002* 1.02
(0.98–1.07)		 0.310 1.01
(0.98–1.04)		 0.728    
LVEF, % 0.99
(0.97–01)		 0.209     1.00
(0.98–1.02)		 0.881    

*P<0.05. CI, confidence interval; HR, hazard ratio; PAF, paroxysmal AF. Other abbreviations as in Table 1.

In the HF group, the cut-off value of RDW for AF recurrence based on ROC curve was 13.9% (area under the curve 0.63, 95% CI 0.51–0.75) with a specificity of 70% and sensitivity of 55%. The prevalence of RDW ≥13.9% was significantly higher in the recurrence group than in the non-recurrence group (23 patients [68%] vs. 11 patients [32%], P=0.025). The distributions of RDW divided into recurrence and non-recurrence groups are shown in Figure 1. An overlap of RDW was found between the recurrence and non-recurrence groups in patients with HF; nevertheless, some cases of recurrence in the HF group had a high RDW value with variability. Moreover, major adverse events occurred in 12 patients (all-cause death in 3 and HF hospitalization in 11) in the HF group. The detailed cause of death was decompensated HF in 2 patients and sepsis in 1. A comparison of the demographic and baseline characteristics of patients in the HF group with and without major adverse events is shown in Table 4. Baseline RDW values were higher in patients with major adverse events than in patients without them (15.0±2.2% vs. 13.8±1.5%, P=0.025). Moreover, the post-ablation AF recurrence rate was higher in patients with major adverse events than in those without them (92% vs. 46%, P=0.004). Subsequent multivariate Cox proportional hazards analysis demonstrated that RDW was an independent predictor of major adverse events in the HF group (HR 1.83, 95% CI 1.13–2.72, P=0.003). The cut-off value of RDW for major adverse events in the HF group based on ROC curve analysis was 14.8% (area under the curve 0.71, 95% CI 0.56–0.86) with a specificity of 85% and sensitivity of 56%. Kaplan-Meier curves demonstrated significant differences within the HF group both in event-free survival for recurrence of AF (RDW ≥13.9% and <13.9%, P=0.019) and for major adverse events (RDW ≥14.8% and <14.8%, P<0.001) (Figure 2).

Figure 1.

Red blood cell distribution width (RDW) distribution divided into recurrence and non-recurrence groups among patients with and without heart failure (HF).

Table 4. Comparison of Demographic and Baseline Characteristics of Patients With and Without Major Adverse Events in the HF Group
  Non-major adverse events
group (n=67)		 Major adverse events
group (n=12)		 P value
Age, years 63.2±11.3 66.1±10.6 0.411
Male sex 50 (75%) 8 (67%) 0.723
BMI, kg/m2 23.9±3.6 22.0±3.9 0.088
Type of AF
　Paroxysmal 21 (31%) 9 (75%) 0.008
　Persistent 36 (54%) 1 (8%) 0.004
　Long-standing persistent 10 (15%) 2 (17%) 0.999
Duration of AF, years 0.9 (0.4–4.5) 0.8 (0.4–7.8) 0.945
Comorbidities
　Hypertension 26 (39%) 4 (33%) 0.999
　Diabetes mellitus 12 (18%) 4 (33%) 0.249
　Coronary artery disease 6 (9%) 2 (17%) 0.600
　Stroke or TIA 3 (5%) 2 (17%) 0.163
Previous device implantation 9 (14%) 4 (33%) 0.103
Laboratory data
　Hemoglobin, g/dl 13.9±2.1 12.7±1.9 0.082
　WBC count (×103 μl) 5.9±1.6 5.7±1.0 0.718
　hs-CRP, mg/L 0.4 (0.80–1.90) 1.6 (0.5–4.2) 0.221
　RDW, % 13.8±1.5 15.0±2.2 0.025
　MCV, fl 92.0±5.6 90.6±6.0 0.422
　Creatinine level, mg/dl 1.0±0.2 1.0±0.3 0.376
　　eGFR, ml/min/1.73 m2 62.3±17.2 55.7±16.3 0.218
　BNP level, pg/dl 127.9 (55.7–252.8) 277.6 (165.5–719.3) 0.004
　Total cholesterol, mg/dl 189.8±35.6 190.8±41.3 0.931
　Albumin, g/dl 4.1±0.4 4.0±0.5 0.273
Echocardiographic data
　LAD, mm 42.5±8.1 45.3±4.6 0.242
　LVEDD, mm 52.4±8.7 59.4±9.6 0.013
　LVEDS, mm 39.2±10.2 48.9±11.3 0.004
　LVEF, % 49.1±14.3 31.3±9.6 <0.001
CHADS2 score 1.8±0.9 2.2±1.6 0.323
CHA2DS2-VASc score 2.6±1.4 3.4±2.0 0.103
NYHA functional class 2.0±0.6 2.5±1.1 0.022
Medical therapy
　ACE-Is or ARBs 38 (57%) 8 (67%) 0.520
　β-blockers 54 (81%) 12 (100%) 0.199
　Spironolactone 39 (58%) 10 (83%) 0.119
　Diuretic 49 (73%) 12 (100%) 0.058
　Digoxin 16 (24%) 1 (8%) 0.445
Ablation procedure
　Pulmonary vein isolation 67 (100%) 12 (100%) NA
　Cavotricuspid isthmus 60 (90%) 10 (83%) 0.620
　LA linear ablation 36 (54%) 7 (58%) 0.768
　CFAE 23 (34%) 5 (42%) 0.745
　Superior vena cava isolation 9 (13%) 0 (0%) 0.341
Recurrence of AF after ablation 31 (46%) 11 (92%) 0.004

Data are presented as number (%), and mean±standard deviation or median (interquartile). Other abbreviations as in Table 1.

Figure 2.

Comparison of the Kaplan-Meier survival curves of the recurrence of atrial fibrillation (AF) (A) and major adverse events (B) after catheter ablation in patients with heart failure (HF) among the 2 groups according to each cut-off RDW value (recurrence of AF: RDW 13.9% and major adverse events: RDW 14.8%). RDW, red blood cell distribution width.

We divided the non-HF group into 3 groups according to 1st and 3rd quartiles of RDW, and compared the outcomes among them (RDW ≥13.6%; n=181, 13.6%>RDW≥12.7%; n=340, and RDW <12.7%; n=157). With regards to the recurrence of AF after ablation, there was no difference in event-free survival among the 3 groups (P=0.585) (Figure 3A). In addition, major adverse events occurred in 5 patients (2 HF hospitalizations, 2 strokes, and 1 death from cancer) among the non-HF patients. Anticoagulant drug therapy was continued at the time of stroke in 1 patient but not in the other because of post-ablation AF non-recurrence. Kaplan-Meier curves demonstrated no significant difference regarding major adverse events among the 3 groups (P=0.874) (Figure 3B).

Figure 3.

Comparison of the Kaplan-Meier survival curves of the recurrence of atrial fibrillation (AF) (A) and major adverse events (B) after catheter ablation in patients with non-heart failure among the 3 groups (RDW ≥13.6%; n=181, 13.6%>RDW≥12.7%; n=340, and RDW <12.7%; n=157). HF, heart failure; RDW, red blood cell distribution width.

Furthermore, the patients in the HF group were divided into impaired LVEF (LVEF ≤40%, n=30) and preserved LVEF (LVEF >40%, n=49) groups. A comparison of the baseline characteristics of the 2 groups is shown in Table 5. Baseline RDW did not differ significantly between the impaired and preserved LVEF groups (14.2±1.7% and 13.9±1.6%, P=0.593). The post-ablation AF recurrence rates were comparable in the 2 groups (50% vs. 55%, P=0.659, respectively). In the preserved LVEF group, multivariate Cox proportional hazards analysis revealed that high RDW exhibited a trend toward an association with AF recurrence after ablation (HR 1.23, 95% CI 0.97–1.54, P=0.082), while the predictive value of RDW was not significant in the impaired LVEF group (HR 0.93, 95% CI 0.61–1.43, P=0.740) (Table S1).

Table 5. Comparison of Demographic and Baseline Characteristics of the Preserved LVEF (LVEF >40%) and Impaired LVEF (LVEF ≤40%) Groups of Patients With HF
  LVEF >40%
(n=49)		 LVEF ≤40%
(n=30)		 P value
Age, years 64.8±10.6 61.8±12.1 0.252
Male sex 34 (69%) 24 (80%) 0.300
BMI, kg/m2 24.4±3.6 22.9±3.6 0.189
Type of AF
　Paroxysmal 19 (39%) 11 (37%) 0.851
　Persistent 24 (49%) 13 (43%) 0.625
　Long-standing persistent 6 (12%) 6 (20%) 0.351
Duration of AF, years 0.8 (0.4–4.5) 0.8 (0.5–8.7) 0.354
Comorbidities
　Hypertension 20 (41%) 10 (33%) 0.506
　Diabetes mellitus 11 (22%) 5 (17%) 0.535
　Coronary artery disease 4 (8%) 4 (13%) 0.470
　Stroke or TIA 2 (4%) 3 (10%) 0.362
Previous device implantation 6 (12%) 7 (23%) 0.223
Laboratory data
　Hemoglobin, g/dl 13.5±1.8 14.1±2.5 0.223
　WBC count (×103 μl) 6.0±1.6 5.7±1.5 0.389
　hs-CRP, mg/L 0.80 (0.40–1.65) 1.00 (0.55–3.18) 0.129
　RDW, % 13.9±1.6 14.2±1.7 0.593
　MCV, fl 91.3±4.4 92.5±7.2 0.392
　Creatinine level, mg/dl 0.9±0.2 1.0±0.3 0.030
　　eGFR, ml/min/1.73 m2 62.7±16.3 59.0±18.4 0.348
　BNP level, pg/dl 115.0 (47.4–208.2) 225.7 (118.5–490.3) 0.003
　Total cholesterol, mg/dl 186.5±33.2 195.5±40.7 0.283
　Albumin, g/dl 4.1±0.4 4.1±0.5 0.880
Echocardiographic data
　LAD, mm 42.2±7.5 44.1±8.1 0.279
　LVEDD, mm 49.6±6.9 59.7±8.9 <0.001
　LVEDS, mm 34.8±6.7 50.4±9.2 <0.001
　LVEF, % 56.2±8.3 30.3±8.2 <0.001
CHADS2 score 1.9±1.1 1.8±1.1 0.732
CHA2DS2-VASc score 2.8±1.5 2.6±1.5 0.565
NYHA functional class 2.0±0.6 2.2±0.8 0.102
Medical therapy
　ACEI or ARB 30 (61%) 16 (53%) 0.490
　β-blockers 40 (82%) 26 (87%) 0.756
　Spironolactone 28 (57%) 21 (70%) 0.253
　Diuretic 36 (74%) 24 (83%) 0.310
　Digoxin 10 (20%) 7 (23%) 0.759
Ablation procedure
　Pulmonary vein isolation 49 (100%) 30 (100%) NA
　Cavotricuspid isthmus 42 (86%) 28 (93%) 0.470
　LA linear ablation 23 (47%) 20 (67%) 0.088
　CFAE 16 (33%) 12 (40%) 0.508
　Superior vena cava isolation 5 (10%) 4 (13%) 0.724

Data are presented as number (%), and mean±standard deviation or median (interquartile). Abbreviations as in Table 1.

We also assessed the RDW value 3 weeks prior to the CA procedure in the outpatient clinic, and the mean RDW values were 13.2±0.8% and 13.9±1.7% in the non-HF (606 patients) and HF (79 patients) groups, respectively. To evaluate the agreement between the RDW values at baseline and 3 weeks before ablation, a Bland-Altman analysis was performed, and the results indicated that the mean differences with 95% limits of agreement were −0.007% (−0.535 to 0.521) and 0.098% (−1.221 to 1.416) in the non-HF and HF groups, respectively (Figure S1). The slopes of the regression line in the Bland-Altman plot indicated no proportional bias in the non-HF (r=0.022, P=0.582) and HF groups (r=−0.062, P=0.585).

Discussion

The present study demonstrated an association between RDW and prognosis after RFCA in patients with AF. High RDW had a significant relationship with recurrence of AF after CA for AF in HF patients, but not in non-HF patients. Furthermore, high RDW predicted major adverse events after CA of AF among the HF patients.

Recent studies have shown a relationship between RDW and mortality among patients with cardiovascular diseases.2–6 Although higher RDW was associated with AF incidence in a population-based cohort and in postoperative patients,7,8 there are limited data regarding the relationship between RDW and prognosis in patients undergoing CA of AF. Gurses et al recently reported that elevated RDW was found to be a significant predictor of late recurrence in 299 patients scheduled for cryoballoon-based AF ablation.9 Another study showed an association between high RDW and AF recurrence after cryoballoon ablation in 49 patients.10 However, data from a large sample for assessing pre-ablation RDW and prognosis after CA of AF do not yet exist. In our assessment of 757 patients with AF divided into HF and non-HF groups, we found a significant relationship between RDW and recurrence of AF after RFCA in HF patients but not in non-HF patients. Our findings provide an additional viewpoint regarding the recurrence of AF after CA for AF patients complicated with HF.

Although the mechanisms underlying the association between higher RDW and poor prognosis are not clearly determined, several possible explanations can be considered in HF patients. Increased activation of the renin-angiotensin system and adrenergic hormones in HF could cause increased RDW with erythropoiesis and reduced cardiac function, resulting in poor prognosis.17 The changes in erythrocyte volume seem to affect the carriage of oxygen to tissues and are associated with free radical and oxidative stress.7,18 This stress also reduces red blood cell survival and could influence RDW. Moreover, inflammation may have an important role in the regulation of RDW by inhibiting the activity of erythropoietin.4,19–21 Several pro-inflammatory cytokines, which are linked to HF, may affect erythropoietin-induced erythrocyte maturation, and decreased erythrocyte maturation could cause a high RDW. Activated inflammatory state has been reported as an important factor in the incidence and maintenance of AF.22 Thus, a possible association between AF and elevated RDW may be considered. All of the abovementioned speculations were based on HF states, and it is plausible that RDW was a prognostic factor in HF patients but not in non-HF patients in the present study. However, inflammatory markers such as hs-CRP and white blood cell count were not significantly associated with AF recurrence in either the HF or non-HF group in our results. The amount of direct inflammatory association between elevated RDW and recurrence of AF after ablation might be weaker than the influence of RDW in HF and other cardiovascular diseases themselves. Moreover, the baseline RDW values in the HF group were widely distributed compared with those in the non-HF group in our study, suggesting that RDW is sufficient to assess prognosis for patients involving 2 cardiovascular pathologies, AF and HF states after CA.

The predictive value of RDW for the recurrence of AF was observed strongly in patients with preserved LVEF compared with those with impaired LVEF. AF is a major etiology in the development of HF with preserved LVEF. It has been hypothesized that some part of the elevated RDW levels at baseline may be influenced by the AF burden in these patients, which could represent the possible association with AF recurrence during the follow-up period. We also found that most of the major adverse events were HF-related, and 92% of the patients with major adverse events had post-ablation AF recurrence, suggesting that RDW is a reasonable prognostic marker in patients with HF and AF. However, for the non-HF group, our finding of an insignificant correlation between RDW and prognosis is contrary to that of a previous study that reported a significant association between elevated RDW and AF recurrence among patients who underwent cryoballoon-based AF ablation.9 The lower prevalence of male patients and persistent AF, lack of excluding HF patients with preserved LVEF (LVEF ≥50%), and small sample size in the former study were considered attributable to the contrasting results. Further large-sample prospective study with adequate power to evaluate RDW as a prognostic marker among non-HF AF patients after catheter ablation is required.

The clinical implication of the present study is that RDW is a potential noninvasive marker for HF patients with an increased recurrence rate following AF ablation. It could also be suggested that high RDW might be useful to guide monitoring (frequency of Holter monitoring or close follow-up) and clinical care (continuation of anticoagulation and antiarrhythmic medications) after CA.

Female sex was another independent predictor of post-ablation AF recurrence in the HF group. The lower efficacy of CA of AF in women and higher incidence of procedural complications than in men have been reported; the female patients with AF had more numerous non-PV foci, meaning that AF was less likely to be completely eliminated; additionally, they had a longer underlying history of AF prior to ablation than the male patients with AF.23,24 Furthermore, women are also reportedly less likely to receive cardioprotective medications such as β-blockers compared with men with chronic HF.25 These factors may have contributed to the poor outcomes of female patients with HF who underwent CA of AF in the present study.

Surprisingly, the traditional prognostic factors of LAD in both the HF and non-HF groups and plasma BNP level in the HF group were not found to be independent predictors for recurrence of AF after ablation therapy in the present study. Simple LAD measurement assessed using 2D M-mode methods sometimes fails to accurately present the whole LA volume.26 Although a substantial rationale for the lack of a predictive LAD value is unclear, the LAD value was not a strong predictor of recurrence, which was overcome by other significant predictors on statistical analyses in this specific study population. Moreover, BNP levels at baseline were relatively widely distributed in the HF patients, so temporal BNP levels may be less likely to predict the recurrence of AF after CA than in lone AF patients.27 It was also unusual that patients with persistent AF rather than paroxysmal AF in the HF group were likely to have fewer episodes of AF recurrence after ablation in the present study. We speculate that, in the HF group, patients with persistent AF were highly considered for selection for CA than those with paroxysmal AF because of the HF condition, and that the more curable persistent AF in HF patients would be subject to ablation therapy. The relatively shorter duration of AF in the HF group than in the non-HF group in this study cohort could support this speculation. Our results from the HF patients were drawn from a small-sample study, which is a potential limitation to evaluating prognosis including well-known prognostic factors.

Study Limitations

This was a retrospective study conducted at a single center. Although we used various monitoring tests to detect the recurrence of AF, asymptomatic short-duration AF may not have been detected in some cases, which could cause underestimation of the recurrence rate during the follow-up period.28 Second, we excluded as much as possible patients with comorbidities likely to be associated with RDW. However, changes of RDW by some unknown underlying disease, or subclinical iron deficiency anemia from anticoagulant drug use, and concealed supplement intakes, which were not found through medical record review, were not completely excluded. Moreover, although differences in RDW between the 3 weeks before ablation and baseline were not significant in the HF and non-HF groups, a concern about RDW stability and variability, especially in patients with HF, which were measured at the specific time point, could exist. Third, stopping amiodarone 1 week before ablation was not enough to remove its effect on atrial electrophysiological properties. Finally, the mean LVEF and New York Heart Association functional class in HF patients indicated that the majority of the patients in the present study had relatively preserved LV function with mild to moderate HF.

Conclusions

In conclusion, high RDW was an independent prognostic marker for the recurrence of AF and major adverse events in patients with HF who underwent RFCA of AF. In contrast, RDW was not associated with recurrence of AF after CA in non-HF patients. In addition to the established factors for recurrence of AF after ablation, RDW is another prognostic marker in AF patients complicated with HF.

Disclosures

No conflict of interest.

Supplementary Files

Supplementary File 1

Table S1. Predictors of baseline characteristics for AF recurrence on univariate and multivariate regression analyses in the LVEF >40% and LVEF ≤40% groups among the HF patients

Figure S1. Bland-Altman plots of difference in RDW between baseline and 3 weeks before ablation against the mean of the 2 measurements in the non-HF group (A), and HF group (B).

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-15-1152

References
 
© 2016 THE JAPANESE CIRCULATION SOCIETY
feedback
 Top