2020 Volume 84 Issue 10 Pages 1709-1717
Background: Atrial fibrillation (AF) recurrence remains a tricky problem in patients undergoing ablation. This meta-analysis aimed to summarize the current literature to clarify whether renin-angiotensin system inhibitors (RASIs) prevent AF recurrence after ablation.
Methods and Results: Relevant studies were searched on Pubmed and EMBASE through December 2019. Pooled relative risk (RR) of AF recurrence was calculated. Subgroup analyses according to study design, race, and follow-up duration were further performed. A total of 15 studies examining 4,300 patients were included, with 3 randomized controlled trials and 12 cohort studies. Overall analysis showed that RASIs significantly reduced AF recurrence after ablation (RR=0.83; 95% confidence interval (CI) 0.70–0.98, P=0.028; I2=68.9%). Subgroup analysis further indicated that positive results were found in randomized controlled trials (RR=0.51, 95% CI 0.37–0.70, P<0.001; I2=4%), studies conducted in Asia (RR=0.59, 95% CI 0.46–0.76, P<0.001; I2=30.7%), and studies with follow-up duration ≥1 year (RR=0.82, 95% CI 0.70–0.95, P=0.01; I2=59.1%); negative results were found in cohort studies, studies conducted in Europe or the USA, and studies with follow-up duration <1 year.
Conclusions: RASIs can potentially prevent AF recurrence after ablation under selected conditions. However, more studies are required to confirm this finding due to the variation in current evidence.
As the most common arrhythmia and characterized by rapid and disordered atrial electrical activity, atrial fibrillation (AF) is considered to be close related to atrial physiological and electrophysiological remodeling.1 The renin-angiotensin system plays an important role in cardiac remodeling, and thus triggers a series of factors causing arrhythmia.2 Previous studies have shown that renin-angiotensin system inhibitors (RASIs), including angiotensin-converting enzyme inhibitors (ACEI) and angiotensin-receptor blockers (ARB), can prevent fibrosis of the left atrium and reverse the process of atrial remodeling.3 Therefore, RASIs are recommended for the prevention and treatment of AF.4
Over the past decade, radiofrequency catheter ablation has been widely used as an effective method of rhythm control in selected AF patients. There is considerable evidence available from prospective randomized controlled trials (RCTs) demonstrating that ablation is superior to antiarrhythmic drugs in controlling AF. However, given the complexity of AF and ablation, AF recurrence remains a tricky problem in some patients undergoing AF ablation. Because of their ability to reverse cardiac remodeling, RASIs might have potential value for preventing AF recurrence after ablation. However, previous studies have yielded contradictory results.5–19 Therefore, an unbiased meta-analysis is needed to establish the clinical role of RASIs. In this study we summarized the currently available evidence about the effect of RASI therapy for the prevention of AF recurrence after ablation, and we discuss the commonalities and differences among these trials.
English-language studies were included in the present meta-analysis if they met all the following criteria: (1) RCT, prospective cohort study, or retrospective cohort study; (2) study population comprised AF patients who underwent radiofrequency ablation; (3) the experimental or exposed group received ACEI or ARB (any dose and any type); and (4) the control or unexposed group received placebo, no treatment, or drugs other than RASIs. Studies were excluded on the basis of the following criteria: (1) case-control and cross-sectional studies; (2) not providing adequate data for calculating relative risk (RR) of endpoints; and (3) possible overlapping of study samples or overt verification bias.
Search Strategy and Study SelectionSearches were performed through the PubMed and EMBASE databases (last search update December, 2019) using the following medical MeSH terms and text words without further restrictions: AF, auricular fibrillation, atrial flutter, catheter ablation, radiofrequency ablation, radio-frequency ablation, ablation, losartan, valsartan, irbesartan, telmisartan, olmesartan, candesartan, sartans, ARB, angiotensin II type 1 receptor antagonists, type 1 ARB, angiotensin 2 type 1 receptor antagonists, selective angiotensin II receptor antagonists, lisinopril, enalapril, fosinopril, cilazapril, benazepril, ramipril, captopril, perindopril, and ACEI. Reference lists of identified articles were also searched. The titles and abstracts of selected articles were analyzed by 2 of the authors (L.P. and Z.L.) independently to evaluate conformity with the inclusion criteria. The full text of an article was scanned carefully if its title and abstract were unclear about the study’s admissibility. Any disagreement between reviewers was resolved by consensus.
Quality AssessmentThe quality of included studies was evaluated independently by 2 reviewers. The quality of RCTs was evaluated in terms of the revised Jadad scale.20 Studies with a score ≥4 were considered as good quality. The quality of cohort studies was evaluated by the Newcastle-Ottawa Scale,21 and studies with a score ≥6 were considered as good quality.
Data ExtractionTwo investigators independently extracted the following data for eligible studies using a standardized form: (1) Information about the studies: author(s), publication date, journal, country, and study design; (2) characteristics of the study population: mean age, sex, number of patients; (3) details of therapy: the dose and type of RASIs, and the ablation procedure; and (4) outcomes: mean follow-up period, the RR of AF recurrence. Disagreements between reviewers regarding data were solved by a third person.
Statistical AnalysisAll statistical analyses were conducted using STATA (version12.0). All P values are 2-tailed. Dichotomous outcome data from the included studies were analyzed by their RRs with 95% confidence interval (CI). Forest plots were used for the graphic representation of data examining heterogeneity, which illustrated the relative strength of treatment effects in multiple quantitative scientific studies in our analysis. Heterogeneity between was analyzed using the I2 test. A P=0.05 was regarded as the critical value for homogeneity. If P<0.05, we used the fixed effect model; otherwise we used the random effects model. Publication bias was assessed by funnel plot and Egger’s test. Meanwhile, sensitivity analysis was performed to investigate if any categorical study characteristic affected study outcomes.
A total of 702 records (98 from Pubmed, 604 from EMBASE) were yielded from the search. After exclusion on the basis of title, abstract and/or text, 15 studies were finally included in the present meta-analysis. Figure 1 showed the flow chart of the article search and selection process.
Flow chart of a meta-analysis of the use of renin-angiotensin system inhibitors to prevent atrial fibrillation recurrence after ablation.
The present meta-analysis involved a total of 4,300 patients in 15 studies. Characteristics of the included studies are summarized in Table 1. There were 3 RCTs and 12 cohort studies, with 6 studies from Asia and 9 from Europe or the USA. The follow-up duration was at least 1 year in 13 studies, and less than 1 year in 2 studies. Most studies adopted radiofrequency catheter ablation, except 1 study that adopted video-assisted thoracoscopic surgical ablation.7 Pulmonary vein isolation was the key procedure of ablation in all studies. In terms of medication, ACEI, ARB or both were used in the experiment groups, whereas statins or blank were used in the control groups.
| Author, year | Journal | Country | Type of catheter ablation |
Follow‐up | No. of Pt |
Study design |
Types of RASIs |
Type of control |
|
|---|---|---|---|---|---|---|---|---|---|
| Wang et al (2016)5 | Heart Rhythm | China | PVI and CTI | 1 year | 256 | RCT | ACEI | Placebo | |
| Mohanty et al (2015)6 |
Heart Rhythm | Italy | PVI | 1 year | 1,048 | PC | ACEI | BLANK | |
| Wang et al (2011)7 | Ann Thorac Surg |
China | VATS, PVI and GPA |
2.2 years | 83 | RCT | ARB | BLANK | |
| Ishikawa et al (2011)8 |
Pacing Clin Electrophysiol |
Japan | PVI and SVCI | 195 days | 264 | RC | ACEI or ARB | BLANK | |
| Klemm et al (2010)9 | Cardiology | Germany | PVI | 2.1 years | 204 | PC | ACEI or ARB | BLANK | |
| Tayebjee et al (2010)10 |
Europace | UK | PVI | 1.7 years | 419 | RC | ACEI or ARB | BLANK | |
| Berkowitsch et al (2010)11 |
Pacing Clin Electrophysiol |
Germany | PVI | 13.8 months | 284 | RC | ACEI or ARB | BLANK | |
| Patel et al (2010)12 | Europace | USA | PVI and SVCI | 2 years | 261 | PC | ACEI or ARB | BLANK | |
| Park et al (2009)13 | Korean Circ J | Korea | PVI and CTI, MVI |
18 months | 152 | RC | ACEI or ARB | BLANK | |
| Al Chekakie et al (2007)14 |
J Cardiovasc Electrophysiology |
USA | PVI and CTI | 13.8 months | 177 | RC | ACEI or ARB | BLANK | |
| Richter et al (2007)15 |
Am Heart J | Austria | PVI | 12.7 months | 234 | RC | ACEI or ARB | Statins or blank |
|
| Anne et al (2004)16 | Heart | Belgium | PVI | 2.2 years | 196 | RC | ACEI or ARB | Other drugs | |
| Cui et al (2015)17 | Exp Ther Med | China | PVI | 13.8 months | 120 | RCT | ARB | BLANK | |
| Rosen et al (2013)18 |
J Atr Fibrillation | USA | PVI and PLAWSD |
135 days | 312 | PC | ACEI or ARB | BLANK | |
| Takigawa et al (2012)19 |
J Renin Angiotensin Aldosterone Syst |
Japan | PVI and CTI | 18.9 months | 292 | RC | ACEI or ARB | BLANK | |
| Author, year | Age, years | Hypertension, n (%) |
HF, n (%) | CAD, n (%) | LAD, mm |
LVEF, % |
Type of AF | AF duration |
Initiation of RASI use |
| Wang et al (2016)5 | 60 | 135 (52.73) | 0 | 22 (8.59) | NA | NA | PAF | NA | Day of ablation |
| Mohanty et al (2015)6 |
62.5 | 631 (60.21) | NA | 175 (16.70) | 44.3 | 52.5 | PAF, nonPAF | NA | At least 3 months before ablation |
| Wang et al (2011)7 | 56 | 37 (44.58) | NA | NA | 52 | 61 | LPeAF | 71 months | NA |
| Ishikawa et al (2011)8 |
63 | 109 (41.29) | NA | 13 (4.92) | NA | NA | PAF, PeAF | 4.2 years | At least 1 month before ablation or immediately after ablation |
| Klemm et al (2010)9 | 63 | 129 (63.24) | NA | 17 (8.33) | 45 | NA | PAF | NA | At least 2 months before ablation |
| Tayebjee et al (2010)10 |
57 | 148 (35.32) | NA | 34 (8.11) | 39 | NA | PAF, nonPAF | 58 months | NA |
| Berkowitsch et al (2010)11 |
61 | 284 (100.00) | NA | 42 (14.79) | NA | 60 | PAF, PeAF | NA | Before ablation |
| Patel et al (2010)12 | 63 | 156 (59.77) | NA | 50 (19.16) | 46.7 | 52 | PAF, PeAF, LPeAF |
NA | At least 3 months before ablation |
| Park et al (2009)13 | 57 | 46 (30.26) | NA | 11 (7.24) | 40.5 | 62.4 | PAF, PeAF | 30 months | At least 1 month before ablation |
| Al Chekakie et al (2007)14 |
56 | 70 (39.55) | NA | NA | NA | NA | PAF, PeAF | 6.06 years | At least 1 month before ablation |
| Richter et al (2007)15 |
56.7 | 128 (54.70) | NA | 33 (14.10) | 45 | 61.3 | PAF, PeAF | 6.1 years | At least 3 months before ablation |
| Anne et al (2004)16 | 57 | 34 (17.35) | 8 (4.08) | 30 (15.31) | 40 | 65 | AFL, AF | 4.6 years | On or before day 0 |
| Cui et al (2015)17 | 63.2 | NA | 0 | 23 (19.17) | NA | 60.6 | PAF, PeAF | 9.1 months | At least 4 weeks beforehand |
| Rosen et al (2013)18 |
64 | 188 (60.26) | 28 (8.97) | 43 (13.78) | NA | 60 | PAF, PeAF, LPeAF | NA | NA |
| Takigawa et al (2012)19 |
61 | 141 (48.29) | NA | NA | 38.2 | 61.5 | PAF, PeAF | 4.43 years | NA |
ACEI, angiotensin-converting enzyme inhibitors; AF, atrial fibrillation; AFL, atrial flutter; BLANK, without drug treatment; CAD, coronary artery disease; CTI, cavotricuspid isthmus ablation; GPA, ganglionic plexus ablation; HF, heart failure; LAD, left atrial diameter; LPeAF, long-lasting persistent AF; LVEF, left ventricular ejection fraction; MVI, mitral valve isthmus ablation; NA, not available; Pt, patient; PAF, paroxysmal AF; PC, prospective cohort; PeAF, persistent AF; PLAWSD, posterior left atrial wall and septal debulking; PVI, pulmonary vein isolation; RASIs, renin-angiotensin system inhibitors; RC, retrospective cohort; RCT, randomized controlled trial; SVCI, superior vena cava isolation; VATS, video-assisted thoracoscopic surgical ablation.
The Jadad scores for the RCTs ranged from 3 to 4 (Table 2). None of the 3 studies had adopted a double-blind design. The Newcastle-Ottawa Scale scores for cohort studies ranged from 6 to 8 (Table 3), indicating good-quality studies.
| Author (year) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mohanty et al (2015)6 |
Ishikawa et al (2011)8 |
Klemm et al (2010)9 |
Tayebjee et al (2010)10 |
Berkowitsch et al (2010)11 |
Patel et al (2010)12 |
Park et al (2009)13 |
Al Chekakie et al (2007)14 |
Richter et al (2007)15 |
Anne et al (2004)16 |
Rosen et al (2013)18 |
Takigawa et al (2012)19 |
|
| Representativeness of exposed cohort |
1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |
| Selection of non- exposed cohort |
1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |
| Ascertainment of exposure |
1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |
| Outcome not present at beginning of study |
1 | 0 | 1 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 0 |
| Controls for basic confounder |
1 | 0 | 1 | 0 | 1 | 1 | 1 | 1 | 1 | 1 | 0 | 0 |
| Controls for additional confounders |
0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 |
| Outcome assessment |
1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |
| Long enough follow-up |
1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |
| Adequacy of follow-up |
1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 0 | 1 |
| Sum | 8 | 6 | 8 | 6 | 7 | 8 | 8 | 7 | 7 | 7 | 6 | 6 |
Individually, 7 studies found that RASIs reduced the postoperative recurrence of AF,5,7–9,11,13,17 but another 7 studies showed a non-statistical association between ACEI/ARB and risk reduction of AF recurrence.6,10,12,14–16,19 Most interestingly, Rosen et al found that RASIs even increased the risk of postoperative recurrence of AF.18 Heterogeneity analysis showed I2=68.9%, so a random effects model was used. Meta-analysis of all included studies showed the pooled RR of AF recurrence in patients receiving RASIs was 0.83 (95% CI 0.70–0.98, P=0.028), indicating that RASIs had statistically significant effects on the prevention of AF recurrence after ablation (Figure 2).
Forest plot of all studies in a meta-analysis of the use of renin-angiotensin system inhibitors to prevent atrial fibrillation recurrence after ablation. CI, confidence interval; RR, relative risk.
Sensitivity analysis showed that our analysis was robust in study selection (Figure 3). The substitution of a random effects model for a fixed model did not change our initial qualitative interpretation. Subgroup analyses according to study design, country, and follow-up duration were subsequently performed.
Sensitivity analysis of all studies in a meta-analysis of the use of renin-angiotensin system inhibitors to prevent atrial fibrillation recurrence after ablation. CI, confidence interval.
Regarding the differences in study design, pooled analysis of the 3 RCTs showed a strong association between RASIs use and risk reduction of AF recurrence (RR=0.51, 95% CI 0.37–0.70, P<0.001), with low heterogeneity among studies (I2=4%). However, pooled analysis of the 12 cohort studies showed no statistical association between ACEI/ARB use and risk reduction of AF recurrence (RR=0.90, 95% CI 0.77–1.06, P=0.214), with significant heterogeneity among studies (I2=64.9%) (Figure 4A).
Forest plots of subgroup analyses with study design (A), country (B) and follow-up duration (C) in a meta-analysis of the use of renin-angiotensin system inhibitors to prevent atrial fibrillation recurrence after ablation. CI, confidence interval; RCT, randomized controlled trial; RR, relative risk.
Regarding the difference in countries, RASI use decreased the risk of AF recurrence in Asian countries (RR=0.59, 95% CI 0.46-0.76, P<0.001; I2=30.7%), but not in Europe or the USA (RR=0.97, 95% CI 0.82–1.16, P=0.752; I2=66.4%) (Figure 4B).
Regarding the difference in follow-up duration, RASI use decreased the risk of AF recurrence in studies with a follow-up duration ≥1 year (RR=0.82, 95% CI 0.70–0.95, P=0.01; I2=59.1%), but not in studies with a follow-up duration <1 year (RR=0.92, 95% CI 0.32–2.63, P=0.876; I2=91.7%) (Figure 4C).
The funnel plots and Egger’s test applied to individual trials did not find significant publication bias (P=0.213; 95% CI −4.46 to1.09; Figure 5).
Funnel plots for studies included in a meta-analysis of the use of renin-angiotensin system inhibitors to prevent atrial fibrillation recurrence after ablation. RR, relative risk.
Radiofrequency catheter ablation is a demonstrated effective treatment for AF in 80% of patients,22 but AF recurrence still occurs in some patients after ablation. Regarding the controversy over RASIs for the prevention of AF recurrence after ablation, the present meta-analysis extends current knowledge about RASI use conferring protection against AF recurrence after ablation. More interestingly, the effects of RASIs vary in patients of different races and with different follow-up duration.
Although the pooled analysis of cohort studies obtained negative results in the present meta-analysis, the great heterogeneity among studies should not be ignored, because it might lower the reliability of the pooled effects. As a more reliable study design, in the present meta-analysis RCTs obtained consistently positive results with less heterogeneity. Therefore, we believe that RASIs can be used to prevent AF recurrence after ablation based on current evidence. However, the benefits of RASIs may vary among different populations. For example, different to the overall trend, Rosen et al found that RASIs had a negative effect on preventing AF recurrence after ablation.18 Compared with the other included studies, patients were older in that study, and more likely to have other underlying disorders including hypertension, coronary artery disease and congestive heart failure. Although these factors were adjusted in the multivariable logistic regression analysis, it is a possible that other important underlying conditions or confounding variables were not fully considered. Besides, in these patients both structural and electrical atrial remodeling, especially fibrosis, were advanced, and at this stage, RASI therapy would not be effective.
According to the results of subgroup analysis, the effect of RASIs in Asian populations was better than in European and American populations. We noticed that patients in the studies of European and American populations were more likely to have other underlying disorders, including hypertension (54.49% vs. 43.43%) and coronary artery disease (13.79% vs. 9.98%), when compared with the studies of Asian populations. These comorbidities may increase the risk of AF recurrence and weaken the role of RASI therapy. In addition, the influence of genetic background on drug efficacy is also worthy of study. A previous study indicated that genetic mutations leading to low renin levels would induce resistance to RASIs in patients with hypertension.23 However, whether there is a racial difference in the effect of RASIs for prevention AF recurrence should be further confirmed. Furthermore, based on the results of this meta-analysis, it is better to use RASI therapy for a longer time under tolerable conditions, which brings more benefits than short-term use. Finally, the use of RASIs might bring extra benefits beyond the prevention effect of AF recurrence, such as reducing the risk of heart failure and stroke. Yusuf et al found that irbesartan significantly reduced the risk of hospitalization for heart failure in patients with AF at a follow-up of 4.1 years (HR 0.86; 95% CI 0.76–0.98; P=0.02).24 Another study showed that losartan significantly reduced the risk of stroke when compared with atenolol in patients with AF (HR 0.55; 95% CI 0.31–0.97; P=0.039).25 Therefore, RASIs are recommended as the cornerstone of treatment for heart failure by many guidelines.26,27 However, more details on RASI administration in AF patients are required in the future. Are there any difference in the efficacy between ACEI and ARB? What is the recommended dose of RASIs? In current clinical practice, the patient’s treatment plan should be based on individual principles.
The potential mechanisms of how RASIs may prevent AF recurrence are as follows: (1) reduction in the diastolic volume load and systolic pressure load of heart; (2) inhibition of fibrosis of the left atrium and thus reversal of cardiac remodeling;28 and (3) inhibition of parasympathetic pathological remodeling. However, previous studies did not find solid evidence to support these hypotheses, and more studies are needed to elucidate the underlying mechanism.
Study LimitationsThe number of included studies and study populations was relatively small. The study quality of the RCTs was acceptable but not perfect enough, because none of the 4 studies had adopted a double-blind design. Doses and type of RASIs were different across the studies, and it is unclear whether that affects the effectiveness of treatment. Finally, gray literature (primarily conference abstracts/presentations, ongoing studies, communication with investigators) was not searched.
In conclusion, the present meta-analysis suggested that RASIs are alternative treatment options for the prevention of AF recurrence after ablation. However, due to the variation in the current evidence, further large-scale prospective RCTs are still needed.
The authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
This work was supported by the National Natural Science Foundation of China (grant no. 81900320), the Medical Science and Technology Research Project of Guangdong Province (grant no. C2019107), basic research funding of Sun Yat-sen University (grant no. 19ykpy40), and the Guangzhou Science and Technology Project (grant no. 201802010048).
None.
The Third Affiliated Hospital of Sun Yat-sen University granted exemption from requiring ethics approval.
The data will be available immediately after the publication, ending 2 years after publication. The data will be shared with anyone on a request basis. The data can be applicable for any kind of analyses, and will be shared as Excel files via E-mail.
