Effects of chronic remote ischemic conditioning on atrial brillation burden in patients with permanent pacemakers

Background: The burden of atrial brillation (AF) has been attractive recently and may be associated with stroke risk and mortality. Remote ischemic conditioning could reduce the incidence, inducibility and sustainability of AF. But the effects of chronic remote ischemic conditioning (CRIC) on atrial brillation burden in patients with pacemaker implanted are still unknown. Design: This was a single ‐ center, prospective, randomized, open-labeled clinical study. Methods: Sixty-six patients with permanent pacemakers were randomly divided into CRIC group and control group after 4 weeks screening. CRIC treatment was performed twice a day for 12 weeks. A remote ischemic conditioning protocol consisted of 4 × 5 min ination/deation of the blood pressure cuff applied in the upper arm to create intermittent arm ischemia. Pacemaker programming was performed for four times: before screening (-4-week), after screening and before randomization (0-week, baseline), 4-week and 12-week follow-up. Data including AF burden, longest duration of AF and cumulative numbers of atrial high-rate episodes (AHREs) were collected. Results: Sixty-one patients (31 patients in CRIC group and 30 patients in control group) completed the study. CRIC was well-tolerated by patients after 12 weeks treatment. The burden of AF in CRIC group was decreased signicantly at 4-week compared with that at 0-week (14.7%±18.5% vs 17.0%±20.7%, p < 0.001), which further decreased at 12-week compared with that at 0-week (8.6%±10.2% vs 17.0%±20.7%, p < 0.001) and that at 4-week (8.6%±10.2% vs 14.7%±18.5%, p < 0.001), which was not observed in the control group. AF burden also reduced signicantly after 12-week CRIC compared with that in control group (8.6%±10.2% vs 17.6%±19.5%, p = 0.013). Repeated measurement ANOVA showed that the changes of AF burden was associated with CRIC instead of time (p < 0.01). In addition, there were trends that longest duration of AF and cumulative numbers of AHREs were reduced after 12-week CRIC. Conclusions: This study suggests that a 12-week course of CRIC treatment could reduce AF burden in patients with permanent pacemakers, supporting widespread use of CRIC in the daily lives of these patients, which need to be veried in the future.


Introduction
Atrial brillation (AF) is the most common persistent arrhythmia, independently associated with a twofold increased risk of all-cause mortality in women and a 1.5-fold in men 1 . AF is also related to higher morbidity of stroke, and prevention of stroke can prolong the patient's life 1 2 . The exact mechanism of AF is very complicated and has not been fully elucidated yet, which may include changes of the extracellular matrix, ion channel, broblast, fat cells, myocyte, autonomic nervous system, etc 1 . AF burden, the percentage of AF lasting time, which is usually detected as atrial high-rate episodes (AHREs) by cardiac implanted electronic devices (CIEDs) 3 , is not a major factor in decision on the choice of an intervention in most guidelines 1 4 . However, some studies recently have shown that AF burden was associated with stroke risk 5 6 and mortality 7 , suggesting that lowering AF burden may decrease the risk of stroke and improve the prognosis.
Remote ischemic conditioning (RIC), induced by repeated, transient and non-lethal ischaemia in a distant tissue, has been attractive as it can provide protection of another organ from damage, which could be generated intra-organ 8 and inter-organ 9 . Animal studies have shown that RIC of different organs such as kindey, liver, mesentery and limbs, could make the heart escaped from ischemia/ reperfusion damage 9 10 .
Some clinical studies followed have con rmed the cardioprotective effects of RIC, which usually consisted of 3 or 4 cycles of intermittent in ation/ de ation of a blood pressure (BP) cuff applied to the arm or leg. It was demonstrated that RIC could improve both short-term 11 and long-term 12 clinical outcomes in patients with ST-segment elevation myocardial infarction. Recently, a meta-analysis pooling the results of 30 trials of coronary artery bypass graft (CABG) or valve surgery found that RIC reduced postoperative troponin release compared to the control arm, with no effects on the incidence of acute kidney injury, acute myocardial infarction, and mortality 13 .
Long-term repeated RIC, now called chronic RIC (CRIC) 14 , may increase the 'dose' of RIC and subsequently lead to enhanced protection 15 . It was showed that RIC before CABG surgery and/or valve surgery could reduce the incidence of postoperative AF by 54% (11% RIC vs 24% control; p = 0.031) 16 . Kosiuk et al recently found RIC reduced the inducibility and sustainability of paroxysmal AF (PAF), which was probably mediated by changes in electrophysiological properties of the atria 17 . But the effect of CRIC on AF burden is still unknown. Our previous study 18 showed that 6-week CRIC could improve heart rate variability (HRV) in mild stable ischemic heart failure, which suggested CRIC may have potential bene ts on heart rhythm by modifying stress of autonomic nervous system 19 . This study aimed to examine whether CRIC can decrease the AF burden or not in patients with CIED.

Ethics Statement
This study was approved by the ethics committee of Zhongda Hospital, Southeast University (Nanjing, China). All the subjects provided a full, informed, written consent before entering the study.

Study population
Patients with dual chamber permanent pacemaker (Medtronic) were recruited (by YJ), who received programming in the outpatient department of Cardiology in Zhongda Hospital since Aug 2017. Patients enrolled before 4-week screening should ful ll the following criteria: (1) pacemaker (DDD, Medtronic Inc., Minnesota, USA) with high sensitivity and speci city for the detection of AF or AHREs has been implanted for at least 3 months; (2) atrial sensitivity was more than 1.5 mV with bipolar sensing; (3) the burden of AF was ≥ 1% and ≤ 70%.
After 4-week screening, the patients should ful ll the following criteria in pacemaker programming before randomization: (1) the burden of AF was ≥ 2% and ≤ 70%; (2) the burden variety of AF was less than 15% compared with that before screening.

Study Design
This was a single-center, prospective, randomized, open-labeled clinical study, and the ow chat is illustrated in Fig. 1. Patients included in the study after 4-week screen were randomly divided into two groups (assigned by CZ) at 1:1 ratio in accordance with the random number table (allocation sequence generated by ZD): 1) Control group, in which patients received regular follow-up without CRIC; 2) CRIC group, in which patients received 12-week CRIC treatment. The baseline data were collected just before randomization. Follow-up visits were scheduled at baseline, 4 weeks, and 12 weeks. CRIC was performed as described elsewhere before 18 . Brie y, a BP cuff was applied to the left upper arm of the patient. BP was measured rstly, and then the BP cuff was in ated to a pressure of 20 mmHg greater than the patient's systolic BP for 5 minutes. Afterwards, the cuff was de ated. Five minutes later, the cuff was in ated and de ated again as before. A RIC protocol applied consisted of 4 × 5 min in ation/ de ation of the cuff. Two physicians con rmed that the subjects operated the RIC procedure correctly in the rst week. CRIC treatment was then repeated each morning and evening mostly by the patients themselves.
According to our preliminary study, we hypothesized AF burden decreased by 10% in CRIC group compared with that in control group, thus 30 patients in each group are necessary to offer 80% power at 0.05 signi cant level according to 1:1 ratio of random grouping. Anticipating 10% of patients potentially non-assessable on 3-month study, 66 randomized patients in total are needed.

Pacemaker programming
All the pacemakers in this study were programmed for at least four times: before screening (-4 week), after screening and before randomization (0-week, baseline), 4-week and 12-week follow-up. Before screening, the algorithm was programmed to optimize AF detection, and the function of atrial tachycardia intervention such as post mode switch overdrive pacing, atrial preference pacing and atrial rate stabilization was turned off, which if the pacemaker offered. The detection threshold for AHREs was set to be 180 beats/min over a minimum of 60 secs. Atrial sensitivity was programmed to 0.5 mV with bipolar sensing. Validation of appropriate detection of AHREs or AF was also carried out by scanning atrial electrograms to exclude false-positive detection. The AF burden which was calculated by the hours of all AHREs divided by the hours of follow-up period, the AF with longest duration, and cumulative numbers of AHREs in each visit programming were recorded.

Study endpoints
The primary endpoint of this study was the change of AF burden after 12-week CRIC. Secondary endpoints were the longest duration of AF and cumulative numbers of AHREs recorded at 4-week and 12week visits.

Statistical analysis
Data were statistically analyzed with SPSS ver.15.0 (SPSS Inc., Chicago, IL, USA). Continuous variables were expressed as the mean ± standard deviation, and categorical variables were expressed as the number and percentage (%). Categorical variables were compared using the χ 2 test, and Fisher's exact probability method was used when T < 5. The normal distribution test was conducted by the Kolmogorov-Smirnov method in the continuous variables, and logarithm transition was done if need. Mean values of clinical characteristics were compared between groups by means of Student's t test,and paired t test were used to compare mean values between baseline and each follow-up for each group. Repeated measurement ANOVA was applied to analyze the effects of time and CRIC on the endpoints of the study including AF burden. A probability value of p < 0.05 was considered signi cant.

Results
A total of 110 patients were screened in this study, and 66 patients (32 males and 34 females) were randomized. CRIC was well tolerated during the 12-week period. At last 61 patients completed the study with 31patients in CRIC group and 30 patients in control. No signi cant difference was observed between the two groups for follow-up lost rate (p = 1.0).

Baseline Characteristics
Mean age of the 61 patients who completed the whole follow-up was 65.0 ± 6.0 years (range: 49-75 years). Totally 72% patients had a history of hypertension, 41% had -coronary artery disease, 13% had Page 6/18 -congestive heart failure, and 21% had diabetes mellitus. The mean AF burden was 17.3%±20.2%, AFepisode was 35.4 ± 27.9 per week, and the longest duration of AF was 19.6 ± 34.0 hours. The rate of atrial pacing was 58.3%±32.4%, and the rate of ventricular pacing was 34.1%±42.8%. All the patient characteristics at baseline did not differ between CRIC group and the control group, as shown in Table 1. Types and doses of cardiac medication in this study were not changed for any of the patients during the 16-week period (including 4-week screening period), and the mean doses of β-blockers, angiotensinconverting enzyme inhibitors or angiotensin receptor blockers, aldosterone receptor antagonists and statins were not different between two groups (Table 1).   Table 3). There were no differences between each visit point in control group (p 0.05), and between the CRIC group and control group at 4-week (p = 0.827) and at 12-week (p = 0.470) ( Table 3). Repeated measurement ANOVA showed that the changes of AF episodes number were associated with CRIC rather than time (p < 0.05, data not shown). Comparison between groups by means of Student's t test, comparison between baseline and each follow-up for each group by means of paired t test, and logarithm transition was done before analysis.

Changes of longest AF duration
The longest AF duration at 12-week visit in CRIC group was signi cantly decreased compared to that at 0week visit (17.8 ± 29.9 hours vs 20.1 ± 31.9 hours, p < 0.001) and that at 4-week visit (17.8 ± 29.9 hours vs 19.1 ± 29.6 hours, p = 0.001), while 4-week CRIC did not reduce the longest AF duration signi cantly (p = 0.052) ( Table 4). There were no signi cant differences of longest AF duration among each visit points in control group. The differences between CRIC group and control group at each visit point were also not signi cant, while repeated measurement ANOVA showed that the changes of longest AF duration were associated with CRIC (p < 0.01, data not shown). Comparison between groups by means of Student's t test, comparison between baseline and each follow-up for each group by means of paired t test, and logarithm transition was done before analysis.

Discussion
The present study has mainly demonstrated that 12-week CRIC can decrease the AF burden in pacemaker-implanted patients, which need to be con rmed in large RCT with long-term follow-up in the future.
AF is a major risk factor for ischemic stroke and provokes important economic burden along with signi cant morbidity and mortality, both incidence and prevalence of which have increased over the last 20 years and will continue to increase over the next 30 years 20 . AF burden had not got enough attention for a long time and incorporated into stroke risks scoring systems may partly because of the di culty to assess these data before. However, with wider usage of different types of CIEDs including pacemakers, ICD, CRT-D/P, and loop recorders, it is easier to monitor long-term atrial arrhythmias and evaluate the AF burden precisely, even if these arrhythmias are asymptomatic and may last only seconds 21 .
Another reason why AF burden was not valued perhaps is the controversial results of stroke risk between PAF and consistent AF in previous studies. Hart et al 22 compared 460 intermittent AF participants with 1,552 sustained AF participants during aspirin therapy in the Stroke Prevention in Atrial Fibrillation studies, and found the annualized rate of ischemic stroke in patients with intermittent (3.2%) was similar to the patients with sustained AF (3.3%) after a mean of two years follow-up. However, another study drew a distinct conclusion 6 . They analyzed 6563 aspirin-treated patients with AF from the ACTIVE-A/AVERROES databases, and found that yearly ischemic stroke rates were 2.1, 3.0, and 4.2% for paroxysmal, persistent, and permanent AF, respectively, with adjusted hazard ratio of 1.83 (P < 0.001) for permanent vs. paroxysmal and 1.44 (P = 0.02) for persistent vs. paroxysmal.
However, AF in those studies was mainly identi ed with infrequent electrocardiographic techniques 22 , which identi ed predominantly PAF who are in AF most of the time. As mentioned above, with the wide application of more sensitive and accessible CIED for AF detection, especially the wide usage of DDD pacemakers, we now recognize a much larger prevalence of paroxysmal and short-lasting AF, for which the association with stroke may be different 3 23 .
Several studies have used CIED to detect AF in different populations and analyzed the association between AF burden and stroke or systemic thromboembolism 3 . Boriani et al performed a pooled analysis of data from 3 prospective studies (PANORAMA, Italian Clinical Services Project, and TRENDS), with 10,016 patients received an CIED implanted, which had at least 3 months of follow-up 24 . After adjustment for CHADS2 score and usage of anticoagulants, Cox regression analysis showed that AF burden was an independent predictor of stroke. The risk for stroke increased by about 3% for every additional hour increase in the daily maximum of AF burden 24 . A recent study 23 found that patients with thromboembolism had higher incidence of pacemaker-detected AF and higher AF burden in the propensity score-matched cohort with comparable CHA2DS2-VASc score, and patients with AF episodes lasting > 6 min had higher risk of future stroke or thromboembolism on Cox regression analysis (propensity-adjusted HR, 6.75; P = 0.023). Besides thromboembolic events, AF burden was associated with the risk of major adverse cardiovascular events (MACE) 25  The results of clinical studies associated with CRIC were not always consistent. A potential reason was the different parameters of CRIC, including pressure, duration of de ation and in ation, number of cycles, frequency of CRIC, etc. The total duration of CRIC also differed greatly in many studies, which may play a role in the divergent results. We found that 12-week CRIC was more effective than 4-week CRIC in reducing AF load analyzed by repeated measurement ANOVA in our study, which means that CRIC may enhance its effects in a dose-dependent patern 15  observed that AF recurrence in the 3 months after the PAF ablation was signi cantly lower in the RIC group than in the control group. The RIPC group also showed a lower increase in atrial remodeling marker Matrix Metalloproteinase-9 and endothelial dysfunction marker von Willebrand-factor in blood samples compared with the control group 30 . Our study has shown that HRV was improved in heart failure patients by CRIC, which was a marker of autonomic nervous stress 19 . As we known, cardiac autonomic nervous stress was closely related to emerging and development of AF. Thus, neural factors may play an important role in CRIC, too. CRIC could regulate immune/ in ammation system 14 and improve cardiac remodeling 31 , which also plays an important role in the progression of AF. However, the exact mechanism in reduction of AF burden by CRIC need to be further exploited in the future.
In conclusion, we found that CRIC can signi cantly reduce the AF load in patients with pacemaker in a dose-dependent manner. CRIC, a noninvasive and promising adjunctive therapy, thereby may further improve the outcomes over well-established therapies in AF patients.
Limitation: The sample size of this study was relatively small, and the course of follow-up was not long, therefore we have not set mortality and morbidity of stroke as end points. The results of our study and other potential meaningful bene ts by CRIC need to be veri ed in large RCT.

Conclusions
A 12-week course of CRIC treatment could reduce AF burden in patients with permanent pacemakers, supporting widespread use of CRIC in the daily lives of these patients, which need to be veri ed in the future. Hospital, Southeast University (Nanjing, China). All the subjects provided a full, informed, written consent before entering the study.
Consent for publication: Not applicable.
Availability of data and materials: The datasets used and/or analysed during the current study are available at http://www.chictr.org.cn/showproj.aspx?proj=21591.
Competing interests: The authors declare that they have no competing interests.
Funding: This work was supported by grant from Nanjing City Development Project of Medical Science and Technology (No. YKK18258), including collection, analysis, and interpretation of data.
Authors' contributions: CL and ZQ designed the study, and CL was a major contributor in writing the manuscript. ZD, YJ and CZ recruited patients, and were responsible for CRIC intervention. YC and LY were in charge of pacemaker programming. JH and RL analyzed and interpreted the patient data. All authors read and approved the nal manuscript.