The correlation of serum long non‐coding RNA ANRIL with risk factors, functional outcome, and prognosis in atrial fibrillation patients with ischemic stroke

Abstract Background This study aimed to evaluate the predictive value of long non‐coding RNA (lncRNA) antisense non‐coding RNA in the INK4 locus (ANRIL) for atrial fibrillation (AF) patients with ischemic stroke and investigate its correlation with risk factors, functional outcome, and prognosis. Methods A total of 386 consecutive AF patients were recruited. AF patients were followed up for 24‐48 months by outpatient follow‐up, telephone follow‐up, and medical record. The time of ischemic stroke in patients with AF was recorded, and follow‐up was continued for 6 months. LncRNA ANRIL expression from serum was detected by quantitative real‐time polymerase chain reaction (qRT‐PCR). Results Compared with the AF with ischemic stroke group (14.3 ± 2.3), patients in the AF without ischemic stroke group (11.9 ± 1.8) had significantly lower serum lncRNA ANRIL levels (P < .05). The sensitivity and specificity of lncRNA ANRIL for identifying AF with ischemic stroke were 76.6% and 81.4%, respectively. Spearman correlation analysis results shown that lncRNA ANRIL was significantly correlated with the NIHSS score (r Spearman = .490, P < .001) and the mRS score (r Spearman = .466, P < .001). Compared with the lncRNA ANRIL high‐expression group, the recurrence‐free survival (RFS) of the lncRNA ANRIL low‐expression group was significantly higher (χ2 = 11.009, log‐rank P < .001). Cox proportional regression model analysis indicated that the serum lncRNA ANRIL level (P = .004), NIHSS score (P = .001), infarct volume (P = .035), and smoking (P < .001) were the risk factors for AF with ischemic stroke. Conclusion Serum lncRNA ANRIL exerts a good predictive value for AF with ischemic stroke, and its increased expression is correlated with worse RFS for patients.


| INTRODUC TI ON
Atrial fibrillation (AF) is the most common persistent arrhythmia. 1 A multi-center study published in 2013 analyzed AF epidemiological data in 10 different regions (4 towns and 6 rural areas) and shown that the prevalence of AF in people aged 35-59 years was 0.42%, and that of those over 60 years old was 1.83%. 2 Among people over 80 years old, the incidence of AF was up to 7.5%. 2 During the onset of AF, the frequency of atrial agitation can reach 300-600 beats per minute, which causes the patient's atrial to lose effective contractile function. 3 The oxidative damage and thrombosis of AF are hot spots in recent years. It involves a variety of oxidative stress injuries including lymphocyte DNA damage and myeloperoxidase, which leads to damage to the brain, heart, kidney, and other organs. [4][5][6] Among them, ischemic stroke is one of the greatest complications of AF, and the consequences of ischemic stroke caused by AF are more serious, and the mortality rate is extremely high. 7,8 Long non-coding RNAs (lncRNAs) are a class of sequences longer than 200 nucleotides that were once considered as junk sequences. 9 The mechanism of lncRNA mainly includes the regulation of transcription and post-transcription. 10,11 Recent studies have shown that ln-cRNA not only affects the progress and outcome of ischemic stroke through ischemia-reperfusion injury, inflammatory response, angiogenesis, nerve repair, and other pathways, and its early diagnosis, disease classification and prognosis assessment are also important. [12][13][14] LncRNA antisense non-coding RNA in the INK4 locus (ANRIL) is the gene most closely linked to atherosclerosis, which is significantly related to the occurrence of cerebral ischemia. 15 In a prospective case-control study, Zhang et al 16  All these studies suggest that lncRNA ANRIL is involved in the process of ischemic stroke. The purpose of this study was to investigate the correlation of serum lncRNA ANRIL in AF with ischemic stroke and to provide a new way for the clinical diagnosis and treatment of AF with ischemic stroke.

| Number of subjects required for current study
The formula for calculating the number of diagnostic test samples is as follows: n = μ 2 ∕ 2 1 − P P, where α is set to 0.05, μ α is set to 1.96, δ is set to .05, and the P Sensitivity is set to 90%. After substituting μ α = 1.96, δ = .10, and P Sensitivity = 90% into the formula, the number of cases that can be obtained is 130; that is, at least 130 patients are required to obtain reliable statistical results in our study.

| Inclusion of study subjects
In this study, patients who did not complete follow-up or were lost to follow-up were excluded (n = 93). In the end, a total of 386 consecutive AF patients who went to the outpatient clinic of The

| Blood samples and basic parameters collection
In the first outpatient visit of 386 patients with AF, 2 mL of peripheral venous blood was collected from each patient using blood collection tubes containing coagulant, and then, the serum was separated and stored in a −80° ultra-low-temperature refrigerator for later use. Moreover, patients' age, gender, body mass index (BMI), hypertension, diabetes, drinking, smoking, blood lipid levels, procalcitonin (PCT), high-sensitivity C-reactive protein (hs-CRP), fibrinogen, D-dimer, family history of ischemic stroke, medications, etc, were collected.
All patients with AF were scored according to the National Institutes of Health Stroke Scale (NIHSS) immediately after the occurrence of ischemic stroke, and the infarct volume was measured by MRI within 24 hours of onset.

| Functional outcome and prognosis evaluation of AF with ischemic stroke at 6 months of follow-up
Functional outcomes, recurrence-free survival, and survival status were assessed at 6 months after ischemic stroke in patients with AF. Functional outcome was evaluated based on a modified Rankin Scale (mRS). An mRS score between 0 and 2 indicates that the patient's functional recovery is good (favorable functional outcome), while an mRS score of 3-6 or death is a poor function recovery (unfavorable functional outcome). In addition, the time to recurrence of ischemic stroke during the 6-month period was recorded.

| Serum total RNA extraction and quantitative real-time polymerase chain reaction (qRT-PCR) detection
Serum total RNA was extracted according to the steps of the serum RNA extraction kit (QIAGEN, Lot number: 50504) and then reversetranscribed into cDNA using the genome-free-reverse transcription The inspection level α in this study was .05.

| Comparison of basic information and lncRNA ANRIL expressions between AF with ischemic stroke group and AF without ischemic stroke group
Of the 386 patients with AF, a total of 132 (34.2%) patients had ischemic stroke. Therefore, patients were divided into AF with ischemic stroke group (n = 132) and AF without ischemic stroke group (n = 254). The basic information of patients is shown in Table 2.
Compared with the AF with ischemic stroke group (14.3 ± 2.3), AF without ischemic stroke (11.9 ± 1.8) had significantly lower serum lncRNA ANRIL levels (P < .05, Figure 1A and Table 2). The qRT-PCR amplification melt curve showed a single amplification product, suggesting high amplification specificity ( Figure 1B). The stability test results are shown in Figure 1C,D. Serum lncRNA ANRIL expression was not affected by blood ex vivo time and storage temperature. In addition, we also found that the number of patients with smoking history in the AF with ischemic stroke group was significantly higher than that in the AF without ischemic stroke group (P < .05; Table 2). There were no significant differences in other parameters between the two groups.

| Risk factors for AF with ischemic stroke
The results of univariate and multivariate logistic regression analyses for AF with ischemic stroke are shown in Table 3. Serum lncRNA ANRIL levels and smoking were independent risk factors for AF with ischemic stroke (P < .05).

| The predictive value of serum lncRNA ANRIL for AF with ischemic stroke
The ROC results are shown in Figure 2. The AUC of serum lncRNA ANRIL for identifying AF with ischemic stroke was 0.826 (0.799-0.873), the cut-off value was 13.2, and the sensitivity and specificity were 76.6% and 81.4%, respectively.

| Correlation of serum lncRNA ANRIL expression with NIHSS score and functional outcome
In order to analyze the correlation between serum lncRNA ANRIL and the severity and functional outcome of patients with ischemic stroke, we analyzed the correlation of serum lncRNA ANRIL expression with NIHSS score and functional outcome. We found that serum lncRNA ANRIL was significantly correlated with the NIHSS score (r Spearman = .490, P < .001; Figure 3A) and the mRS score (r Spearman = .466, P < .001; Figure 3B). The odds ratio of multivariate analysis was adjusted for all significant outcome predictors of the univariate logistic regression analysis.

F I G U R E 2
Receiver operating characteristic (ROC) curve analysis for the differential predictive value of serum lncRNA ANRIL for AF with ischemic stroke

| Correlation of serum lncRNA ANRIL expression with prognosis
A total of 74 patients had recurrent ischemic stroke, and 48 patients died in this study. AF with ischemic stroke patients was divided into the lncRNA ANRIL high-expression group and the lncRNA ANRIL low-expression group according to the median serum ANRIL level.
The results of Cox proportional regression model analysis of risk factors for recurrence-free survival of AF with ischemic stroke are shown in Table 4. The serum lncRNA ANRIL level (P = .004), NIHSS score (P = .001), infarct volume (P = .035), and smoking (P < .001) were the risk factors for AF with ischemic stroke (Table 4). LncRNA ANRIL is the best replicated genetic risk factor for coronary artery disease and regulates genes involved in fatty acid and glucose metabolism. 20 Previous studies showed that ANRIL was highly expressed in endothelial and may be involved in the thrombogenesis. 21,22 Moreover, increased ANRIL has long been investigated as a potential marker in the plaque of atherosclerosis patients. 23   Malonaldehyde produces cytotoxicity by cross-linking polymerized proteins, nucleic acids, and other macromolecules, and promotes endothelial cell damage. 26 It is worth noting that we also found that compared with the lncRNA ANRIL high-expression group, the recurrence-free survival of the lncRNA ANRIL low-expression group was significantly higher. Moreover, the serum lncRNA ANRIL level (P = .004), NIHSS score (P = .001), infarct volume (P = .035), and smoking (P < .001) were the risk factors for AF with ischemic stroke. Abbreviations: BMI, body mass index; CI, confidence interval; HDL-C, high-density lipoproteincholesterol; HR, hazard ratio; Hs-CRP, high-sensitivity C-reactive protein; LDL-C, low-density lipoprotein-cholesterol; Lp (a), lipoprotein (a); mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; PCT, procalcitonin; SE, standard error; TC, total cholesterol; β, regression coefficients.

| D ISCUSS I ON
TA B L E 4 Cox proportional regression model analysis of risk factors for AF with ischemic stroke The above results further confirm that serum lncRNA ANRIL can promote ischemic stroke in patients with AF and is closely related to patients' prognosis.
In this study, we found that serum lncRNA ANRIL has potential diagnostic and prognostic value in AF with ischemic stroke and is expected to provide new approaches and ideas for the diagnosis and treatment for AF patients.

AUTH O R CO NTR I B UTI O N S
JJ conceived the study. WZ and JJ were involved in gaining ethical approval, patient recruitment, and data analysis. WZ wrote the first draft of the article. All authors reviewed and edited the article and approved the final version of the article.

E TH I C A L A PPROVA L
This study was approved by the Ethics Committee of The University of Hong Kong-Shenzhen Hospital ethics committee (LH-S2015073).