LOXL2 as a Predictive Biomarker for Breakthrough Stroke Risk in Paroxysmal Atrial Fibrillation Patients

LOXL2 as a Predictive Biomarker for Breakthrough Stroke Risk in Paroxysmal Atrial Fibrillation Patients | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 4 March 2025 V1 Latest version Share on LOXL2 as a Predictive Biomarker for Breakthrough Stroke Risk in Paroxysmal Atrial Fibrillation Patients Authors : Fang Liu 0000-0002-4528-6087 , Zhitian Li , Ruolin Wang , Shanshan Meng , Ziting Wu , Dongtao Zhou , Qianran Luan , … Show All … , Ying Dong , Chen-Xi Jiang , Ribo Tang , Wei Wang , Xin Zhao 0000-0003-3663-6097 , Changyi Li , Tong Liu 0000-0001-5372-7236 , Yue-Xin Jiang , Mengmeng Li , Deyong Long 0000-0003-4604-5346 , and Yuanfeng Gao [email protected] Show Fewer Authors Info & Affiliations https://doi.org/10.22541/au.174110600.06008991/v1 219 views 180 downloads Contents Abstract Supplementary Material Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Background: In atrial fibrillation (AF) patients, breakthrough stroke is not uncommon and represents an important subgroup due to its high stroke recurrence rate and mortality. However, no reliable tool exists for assessing stroke risk in anticoagulated paroxysmal AF (PAF) patients. While LOXL2 is implicated in atrial fibrosis, a key pathological substrate for atrial thrombus, its predictive value for stroke remains unclear. Objective: To investigate predictive value of serum LOXL2 levels for breakthrough stroke risk in PAF patients. Methods: We consecutively enrolled 197 anticoagulated PAF patients. The serum level of LOXL2 were quantified via ELISA. Patients were stratified into LOXL2+ and LOXL2− groups based on the median of their baseline LOXL2 (275.9 pg/mL). Stroke events were recorded over a median follow-up of 3.9 years. Predictive models incorporating LOXL2, age, and CHA2DS2-VASc were evaluated using ROC, NRI, and DCA. Results： During follow-up, 24 patients (12.2%) experienced stroke. LOXL2 levels were significantly higher in stroke cases (P = 0.006). Multivariable Cox analysis identified LOXL2 as an independent risk factor (P < 0.001). Kaplan-Meier and Nelson-Aalen cumulative hazard analyses further confirmed the contribution of LOXL2 for elevated stroke risk. A prediction model, incorporating both LOXL2 and age, achieved the highest predictive accuracy (AUC = 0.842), significantly improving risk stratification over CHA2DS2-VASc (NRI = 15.0%, P < 0.001). Conclusion: Elevated LOXL2 is independently associated with breakthrough stroke risk in PAF patients. Incorporating LOXL2 and age enhances predictive accuracy, offering a novel tool for personalized stroke risk stratification in AF patients despite anticoagulation medication. LOXL2 as a Predictive Biomarker for Breakthrough Stroke Risk in Paroxysmal Atrial Fibrillation Patients Short title: LOXL2 and Breakthrough Stroke Risk in PAF Patients Fang Liu, MD PhD a† , Zhitian Li, MD b† , Ruolin Wang, MD a† , Shanshan Meng, BS c , Ziting Wu, PhD c , Dongtao Zhou, MD a , Qianran Luan, MS e , Ying Dong, PhD f , Chenxi Jiang, MD PhD a , Ribo Tang, MD PhD a , Wei Wang, MD PhD a , Xin Zhao, MD PhD a , Changyi Li, MD PhD a , Tong Liu, MD PhD a , Yuexin Jiang, MD PhD a , Mengmeng Li, MD PhD a , Deyong Long, MD PhD a* , Yuanfeng Gao, MD PhD f* , a Department of Cardiology, Beijing Anzhen Hospital, Capital Medical University, National Clinical Research Center for Cardiovascular Diseases, Beijing, China. b Department of Thoracic Surgery, Beijing Shijitan Hospital Affiliated to Capital Medical University, Beijing, China. c North Ward, the Third Hospital of Peking University, No.10 Lanlangou, Haidian District, Beijing, China. d Institute for Global Health and Development, Peking University, Beijing, 100871, China; China Center for Health Economic Research, Peking University, Beijing, 100871, China; National School of Development, Peking University, Beijing, 100871, China. e Department of gerontology The Third People Hospital of Chengdu, Sichuan, China, f Department of Cardiology, Beijing Chaoyang Hospital, Capital Medical University, Beijing, China. * Corresponding author & Permanent address: Deyong Long, Department of Cardiology, Beijing Anzhen Hospital, Capital Medical University, National Clinical Research Center for Cardiovascular Diseases, Beijing, China. E-mail: [email protected] . Yuanfeng Gao, Heart Center & Beijing Key Laboratory of Hypertension, Beijing Chaoyang Hospital, Capital Medical University, Beijing, China, Tel: +86 01085231937, E-mail: [email protected] †: These authors contributed equally to this work. Conflict of interest statement: The authors declare no conflicts of interest. Word count: text proper: 2717 words; total word count: 5285; the number of references: 22. ABSTRACT Background: In atrial fibrillation (AF) patients, breakthrough stroke is not uncommon and represents an important subgroup due to its high stroke recurrence rate and mortality. However, no reliable tool exists for assessing stroke risk in anticoagulated paroxysmal AF (PAF) patients. While LOXL2 is implicated in atrial fibrosis, a key pathological substrate for atrial thrombus, its predictive value for stroke remains unclear. Objective: To investigate predictive value of serum LOXL2 levels for breakthrough stroke risk in PAF patients. Methods: We consecutively enrolled 197 anticoagulated PAF patients. The serum level of LOXL2 were quantified via ELISA. Patients were stratified into LOXL2+ and LOXL2− groups based on the median of their baseline LOXL2 (275.9 pg/mL). Stroke events were recorded over a median follow-up of 3.9 years. Predictive models incorporating LOXL2, age, and CHA2DS2-VASc were evaluated using ROC, NRI, and DCA. Results： During follow-up, 24 patients (12.2%) experienced stroke. LOXL2 levels were significantly higher in stroke cases (P = 0.006). Multivariable Cox analysis identified LOXL2 as an independent risk factor (P < 0.001). Kaplan-Meier and Nelson-Aalen cumulative hazard analyses further confirmed the contribution of LOXL2 for elevated stroke risk. A prediction model, incorporating both LOXL2 and age, achieved the highest predictive accuracy (AUC = 0.842), significantly improving risk stratification over CHA2DS2-VASc (NRI = 15.0%, P < 0.001). Conclusion: Elevated LOXL2 is independently associated with breakthrough stroke risk in PAF patients. Incorporating LOXL2 and age enhances predictive accuracy, offering a novel tool for personalized stroke risk stratification in AF patients despite anticoagulation medication. Keywords ：Paroxysmal Atrial Fibrillation, Breakthrough Stroke, Biomarker, LOXL2, Predictive model ABBREVIATIONS AND ACRONYMS: AF = atrial fibrillation BNP = B-type natriuretic peptide ELISA = enzyme-linked immunosorbent assay LOXL2 = lysyl oxidase like 2 RFCA = radiofrequency ablation OAC = oral anticoagulant TnI = troponin I Introduction: Stroke ranks as the second leading cause of death globally and is a leading contributor to acquired disability in adults. 1 Atrial fibrillation (AF), the most common cardiac arrhythmia, affects nearly 4% of the adult population in high-income countries and serves as a major risk factor for cardioembolic stroke, accounting for approximately 2 one-quarter of all ischemic strokes. While the widespread use of oral anticoagulants (OACs) has significantly reduced ischemic stroke incidence in AF patients, strokes still occur (1–2% per year) 3-5 in some individuals despite anticoagulation therapy. Traditionally, stroke risk in patients with AF has been evaluated using clinical risk scores such as the CHA2DS2-VASc score, which guides anticoagulation decisions. However, the ABC-stroke score, introduced in 2016, has demonstrated superior performance by incorporating biomarkers alongside clinical variables. 6 This has fueled growing interest in incorporating biomarkers into stroke risk assessment for AF patients. The evaluation of thrombogenesis-related biomarkers in AF could pave the way for a more precise and personalized approach to stratify stroke risk. 7, 8 Atrial remodeling, marked by structural and functional alterations in atrial tissue, is increasingly recognized as a critical contributor to arrhythmogenesis. Atrial cardiomyopathy, which embodies these remodeling processes, is now believed to play a central role in facilitating atrial fibrillation (AF) and promoting thrombogenesis. The limited efficacy of rhythm-control strategies in reducing stroke risk, coupled with the lack of a consistent temporal association between paroxysmal AF episodes and stroke occurrence, underscores the potential role of fibrotic atrial cardiomyopathy in thromboembolism. Although the mechanisms linking AF and stroke remain incompletely understood, biomarkers of atrial myopathy may provide valuable insights, offering the potential to refine stroke risk stratification in patients with AF. Lysyl Oxidase Like 2 (LOXL2), a critical mediator in cardiac remodeling, plays a pivotal role in the cross-linking of collagen within the extracellular matrix, thereby influencing the mechanical properties and structural stability of the heart. 9 Beyond this, LOXL2 is involved in the trans-differentiation of fibroblasts into myofibroblasts, a key process in atrial remodeling. 10 Elevated LOXL2 levels have been detected in the serum of patients with atrial AF and in models of Angiotensin II-induced atrial fibrosis. 11, 12 Furthermore, research indicates that LOXL2 inhibition may mitigate atrial fibrosis and reduce AF susceptibility. For instance, the selective LOXL2 inhibitor LOXL2-IN-1 has been shown to significantly lower the vulnerability to AF induced by Ang II, while also reducing atrial inflammation, fibrosis, and cardiac hypertrophy. However, to date, no studies have comprehensively assessed the predictive capacity of serum LOXL2 levels in relation to stroke risk in patients with AF receiving oral anticoagulation therapy. This study aims to explore the association between baseline LOXL2 levels and breakthrough stroke risk in paroxysmal atrial fibrillation (PAF) patients. By evaluating the added predictive value of LOXL2 to the CHA2DS2-VASc score, this research seeks to establish a stronger link between cardiac remodeling and stroke occurrence, positioning LOXL2 as a potential biomarker for improved risk stratification. Study Design and Participants This prospective cohort study enrolled 197 individuals diagnosed with symptomatic non-valvular PAF. Patients were excluded if they had a history of acute or chronic inflammation, myocardial infarction or stroke within the preceding three months, thyroid dysfunction, congenital heart defects, recent cardiac surgery (within three months), or severe valvular heart disease. The study protocol was approved by the Ethics Committee of Beijing Chaoyang Hospital, Capital Medical University, China (2024-ke-23). Written informed consent was obtained from all participants, and the study adhered to the principles outlined in the Declaration of Helsinki. Laboratory Measurements and LOXL2 Grouping Fasting blood samples were obtained between 6:00 and 7:00 a.m. on the second day post-admission. Serum levels of LOXL2, along with B-type natriuretic peptide (BNP), troponin I (TnI), and C-reactive protein (CRP), were quantified using enzyme-linked immunosorbent assay (ELISA). Plasma was separated via centrifugation at 3000 g for 15 minutes. LOXL2 levels were measured using a commercially available ELISA kit (Human LOXL2 ELISA Kit, ab213808, Abcam), following the manufacturer’s instructions. Standard curves were generated using serial dilutions of recombinant human LOXL2, and serum concentrations were determined by extrapolation using a log–log linear regression model. Other biochemical parameters were assessed at the central laboratories of Beijing Chaoyang Hospital. Patients were classified as LOXL2+ or LOXL2− based on the median LOXL2 level among all patients. A cutoff value of 275.9 pg/mL was determined, yielding a sensitivity of 70.8% and specificity of 53.2%. Catheter Ablation and Echocardiographic Assessment During RFCA, pulmonary vein isolation was confirmed in all patients. If atrial fibrillation persisted or reappeared following isolation, additional ablation was performed at the discretion of the operator. Electrical cardioversion was applied in cases of sustained atrial arrhythmias post-ablation. Comprehensive echocardiographic evaluations were conducted within three days of admission using a GE Healthcare ultrasound system (Connecticut, USA). Follow-Up and Stroke Definitions Patients were followed regularly post-discharge. Baseline and follow-up assessments included 12-lead electrocardiograms (ECG) or 24-hour Holter monitoring at 1, 3, 6, and 12 months, and annually thereafter. Antiarrhythmic drugs were usually continued for 3 months post-ablation and systematically discontinued thereafter. More frequent follow-up visits were scheduled as needed for patients with any symptoms indicative of potential stroke. A new onset stroke is defined as detecting a novel lesion via computed tomography (CT) or magnetic resonance imaging (MRI) or the presence of clinical manifestations indicative of a stroke persisting beyond a 24-hour period. 13 Patients were divided into two groups during the follow-up period: those with atrial fibrillation who experienced an ischemic stroke (n=24) and those with atrial fibrillation but no stroke (n=173). Statistical Analysis Continuous data were presented as mean ± standard deviation (SD) or median with interquartile range (IQR), depending on distribution normality. Categorical variables were expressed as percentages. Between-group differences were assessed using independent t-tests, Mann-Whitney U tests, or chi-squared tests, as appropriate. The correlations between these variables were evaluated using the Spearman rank correlation coefficient. Statistical significance was determined with a p-value < 0.05. Multivariable logistic regression and Cox proportional hazards models were used to identify predictors of outcomes, with variables selected via stepwise procedures. Kaplan-Meier curves were plotted to assess time to stroke, with statistical significance evaluated using log-rank tests. The added predictive value of LOXL2 was assessed using Harrell’s C-statistic, net reclassification improvement (NRI), and decision curve analysis (DCA). Receiver operating characteristic (ROC) curve analysis was conducted to determine the area under the curve (AUC), sensitivity, and specificity. All statistical analyses were performed using R software (version 4.4.1, R Foundation for Statistical Computing, Vienna, Austria). A two-sided P-value < 0.05 was considered statistically significant. Results: LOXL2 Was Significantly Elevated in PAF Patients with Breakthrough Stroke During Follow-Up A total of 197 patients with paroxysmal atrial fibrillation were enrolled, with a mean age of 67.3 ± 10.1 years, and 113 (57.4%) were male. Of these, 150 (76.1%) underwent radiofrequency catheter ablation during hospitalization. At the time of admission, 85 patients (43.1%) presented with atrial fibrillation rhythm; however, following treatment, only 8 (4.1%) remained in AF at discharge. A prior history of stroke was documented in 20 (10.2%) patients. Stroke risk assessment using the CHA2DS2VASc score classified 141 patients (71.6%) as high risk (≥2 in men, ≥3 in women). Notably, five patients (2.5%) did not receive OACs therapy throughout the follow-up period, as their CHA2DS2VASc scores were below 2 (four patients had a score of 1, and one had a score of 0). During a follow-up period of up to 6.48 years, with a median follow-up time of 3.88 years, 24 patients (12.2%) had breakthrough stroke. The demographic, clinical, and biochemical characteristics of the study population, stratified by the presence of incident stroke, are summarized in Table 1 . Patients who experienced stroke were significantly older compared to those without (74.5 ± 8.0 vs. 66.3 ± 9.9, p < 0.001). Additionally, a higher proportion of stroke patients presented with atrial fibrillation rhythm at the time of admission (16.7% vs. 2.3%, p = 0.005). Comorbidities, including a history of stroke (p = 0.027), hyperlipidemia (p = 0.049), carotid artery disease (p < 0.001), and renal dysfunction (p = 0.029), were more prevalent among the stroke group. Furthermore, a lower proportion of stroke patients underwent radiofrequency catheter ablation during their hospitalization, with only 54.2% of stroke patients receiving RFCA compared to 79.2% of those without stroke (p = 0.015). Regarding the CHA2DS2VASc score, there were no low-risk patients (defined as male patients with a score of 0, and female patients with a score of 0 or 1). Among the cohort, 91.7% of breakthrough stroke patients were classified as high-risk according to the CHA2DS2VASc criteria ( Figure 1A ). Among the five patients who did not receive anticoagulation, one patient, with a CHA2DS2-VASc score of 1, experienced a stroke ( Figure 1B ). Moreover, serum LOXL2 levels were significantly higher in stroke cases compared to non-stroke cases (419.8 ± 229.4 vs. 278.8 ± 81.4, P = 0.006, Figure 1C ). However, no significant correlation was observed between LOXL2 levels and BNP (Spearman r = 0.13, P = 0.06), TNI (Spearman r = -0.08, P = 0.24), or CHA2DS2-VASc score (Spearman r = 0.06, P = 0.38) ( Figure 1D ). LOXL2 is an Independent Risk Factor for Breakthrough Stroke in Patients with PAF To identify independent risk factors for stroke, we first conducted univariate analysis and then incorporated all variables with a p-value less than 0.05 into a multivariate Cox regression model. The results indicated that older age (HR = 1.13, 95% CI = 2.80-3.47, p = 0.012), use of antiplatelet medication (HR = 3.21, 95% CI = 2.92-14531, p = 0.037), and elevated serum LOXL2 levels (HR = 1.01, 95% CI = 2.93-2.75, p < 0.001) were independent predictors for incident stroke in patients with PAF( Table 2, Figure 2A ). Taken together, these findings indicate that elevated LOXL2 level could be an independent risk factor for stroke, with higher LOXL2 levels being associated with an increased risk of stroke in these patients. LOXL2 as an Additional Biomarker for the Prediction of Breakthrough Stroke in Patients with PAF To investigated the predictive value of LOXL2 in breakthrough stroke incidences, we first classified patients into two groups based on the median LOXL2 level (275.9 pg/mL). Among the 197 patients, 99 (50.3%) were categorized as LOXL2- and 98 (49.7%) as LOXL2+. The demographic, clinical, and biochemical characteristics of the study population, categorized by the occurrence of incident stroke, are presented in Table 3 . There were no significant differences were found between the two groups, except for LOXL2 levels. In the LOXL2+ group, the proportion of patients who underwent stroke (17/98, 17.3%) were significantly higher (p = 0.047) than that in the LOXL2- group (7/99, 7.1%) ( Figure 2B ). Next, to further clarify the predictive value of LOXL2, we performed Kaplan-Meier survival analysis and Nelson-Aalen cumulative hazard curve analysis. The Kaplan-Meier analysis revealed a significant difference in cumulative survival between the LOXL2+ and LOXL2- groups (p = 0.030, Chi-square value: 4.721). Additionally, the Nelson-Aalen cumulative hazard curve analysis demonstrated a higher cumulative risk in the LOXL2+ group, with a cumulative hazard ratio of 2.56 compared to the LOXL2- group ( Figure 2C, D ). To evaluate the predictive capability of LOXL2 compared to the CHA2DS2-VASc score and to develop a more effective model for predicting breakthrough stroke, we constructed four predictive models. Model A was based solely on the CHA2DS2-VASc score, while Model B incorporated LOXL2 levels. Model C combined LOXL2 and age, whereas Model D integrated LOXL2 with the CHA2DS2-VASc score. The performance of these models was assessed using receiver operating characteristic (ROC) curves, as illustrated in Figure 3A . Notably, Model C, which incorporated both LOXL2 and age, demonstrated the highest predictive accuracy (AUC: 0.842, 95% CI: 0.761–0.923). To further evaluate Model C from multiple perspectives, we conducted additional analyses. First, we assessed the temporal stability of these models using dynamic ROC analysis ( Figure 3B ). Model C exhibited the most consistent predictive performance over time, with only a slight decrease in AUC during the follow-up period. Additionally, to quantify the improvement of Model C over Model A, we performed decision curve analysis (DCA) ( Figure 3C ) and calculated the net reclassification improvement (NRI) ( Table 4 ). The results demonstrated that Model C significantly enhanced stroke prediction by 15% compared to Model A (NRI: 15.0%, 95% CI: 0.10–0.20, P < 0.001). Discussion: This is the first study to identify serum LOXL2 as a potential biomarker for breakthrough stroke risk prediction in patients with PAF receiving oral anticoagulation therapy. We find that LOXL2 levels are significantly elevated in patients who experienced breakthrough stroke, and could be an independent risk factor for breakthrough stroke events in PAF patients. Notably, LOXL2 provides unique predictive value beyond conventional clinical risk models. These findings underscore that LOXL2 as a potential biomarker for enhancing risk stratification of breakthrough strokes and optimizing therapeutic strategies for secondary stroke prevention in PAF patients. Atrial remodeling, particularly atrial fibrosis, is a well-recognized driver of stroke risk in AF patients. Fibrotic remodeling alters atrial structure and function, promoting atrial thrombus formation through abnormal hemodynamics and endothelial dysfunction. 14,15 The lysyl oxidase (LOX) enzyme family, including LOXL2, plays a crucial role in fibrotic processes across multiple organ systems, including the cardiovascular systems. 16-20 Elevated serum LOXL2 levels have also been associated with heart remodeling processes such as heart failure and hypertrophic cardiomyopathy. 21, 22 Emerging evidence suggests that LOXL2 directly contributes to atrial fibrosis, with Zhang et al. demonstrating that LOXL2 is upregulated in Angiotensin II-induced atrial fibrosis and that its inhibition mitigates atrial fibrosis and AF susceptibility. 21 Similarly, Wang et al. reported a positive correlation between LOXL2 levels and atrial fibrosis severity in AF patients. 22 Our findings align with and extend previous research on stroke risk prediction in AF patients. The elevated expression of LOXL2 may be closely linked to the progression of atrial fibrosis and atrial cardiomyopathy. While prior studies have established LOXL2’s role in atrial, its relationship with thromboembolic risk had not been investigated. Our study bridges this gap by providing the first clinical evidence that elevated LOXL2 is independently associated with stroke in anticoagulated AF patients. The CHA2DS2-VASc score is a widely used risk stratification tool for determining the need for anticoagulation in AF patients to prevent stroke. However, there is currently no established risk score to assess the likelihood of breakthrough stroke despite anticoagulation. Our study aims to address this gap by identifying patients who remain at risk for stroke even after anticoagulation therapy. Conventional risk models primarily incorporate clinical variables but may not fully capture key pathophysiological mechanisms, such as atrial fibrosis and thrombogenesis. Our findings indicate that LOXL2 could enhance stroke risk assessment by offering complementary biological insights beyond established scoring systems. This distinction has important clinical implications, as it paves the way for improving risk prediction strategies. Integrating LOXL2 into predictive models may refine risk stratification, identifying high-risk patients who might benefit from closer monitoring or alternative therapeutic strategies. Furthermore, our study provides a foundation for future research, including the development of more precise stroke risk models for anticoagulated AF patients and the exploration of LOXL2-targeted therapies as a potential approach to mitigating atrial fibrosis and reducing stroke incidence. Limitations Despite the promising results, our study has several limitations. First, the modest sample size (n=197) may limit the generalizability of our findings. Nevertheless, we employed rigorous statistical methodologies to ensure robust internal validity and mitigate potential bias, thereby supporting the reliability of our conclusions. Second, while we identified a significant association between LOXL2 levels and stroke risk, this observational design precludes causal inference. As emphasized in the introduction, LOXL2’s mechanistic involvement in stroke pathogenesis likely involves multifaceted pathways. Rather than investigating molecular mechanisms, our primary focus was to establish its clinical relevance as a mediator of atrial remodeling and thromboembolic risk in anticoagulated patients. Importantly, this work provides a foundation for future mechanistic studies to unravel LOXL2’s precise role in cardiac thrombus formation, particularly through targeted investigations of its interactions with extracellular matrix components and coagulation cascades. Conclusion Our study provides preliminary evidence supporting LOXL2 as a novel biomarker for stroke risk prediction in PAF patients receiving anticoagulation therapy. Funding This work was supported by Beijing Nova Program (grant number: 2021041). Data Availability The data used to support the findings of this study are included in the article. Conflict of Interest All authors have reported that they have no relationships relevant to the contents of this paper to disclose. Informed Consent Statement Informed consent was obtained from all the subjects involved in the study. References: 1. Donnan, G.A., M. Fisher, M. Macleod, and S.M. Davis, Stroke. Lancet, 2008. 371 (9624): p. 1612-23.2. Kornej, J., C.S. Börschel, E.J. Benjamin, and R.B. Schnabel, Epidemiology of Atrial Fibrillation in the 21st Century: Novel Methods and New Insights. Circ Res, 2020. 127(1): p. 4-20.3. Connolly, S.J., Ezekowitz, M.D., Yusuf, S., et al. Dabigatran versus warfarin in patients with atrial fibrillation. N. Engl. J. Med. 2009, 361, 1139–1151.4. Giugliano, R.P., Ruff, C.T., Braunwald, E., et al. Edoxaban versus warfarin in patients with atrial fibrillation. N. Engl. J. Med. 2013, 369, 2093–2104.5. Granger, C.B., Alexander, J.H., McMurray, J.J., et al. Apixaban versus warfarin in patients with atrial fibrillation. N. Engl. J. Med. 2011, 365, 981–992.6. Hijazi, Z., J. Lindbäck, J.H. Alexander, et al., The ABC (age, biomarkers, clinical history) stroke risk score: a biomarker-based risk score for predicting stroke in atrial fibrillation. Eur Heart J, 2016. 37(20): p. 1582-90.7. Goldberger, J.J., R. Arora, D. Green, et al., Evaluating the Atrial Myopathy Underlying Atrial Fibrillation: Identifying the Arrhythmogenic and Thrombogenic Substrate. Circulation, 2015. 132(4): p. 278-91.8. Calenda, B.W., V. Fuster, J.L. Halperin, and C.B. Granger, Stroke risk assessment in atrial fibrillation: risk factors and markers of atrial myopathy. Nat Rev Cardiol, 2016. 13(9): p. 549-59.9. Al-U’datt, D., B.G. Allen, and S. Nattel, Role of the lysyl oxidase enzyme family in cardiac function and disease. Cardiovasc Res, 2019. 115(13): p. 1820-1837.10. Adam, O., D. Lavall, K. Theobald, et al., Rac1-induced connective tissue growth factor regulates connexin 43 and N-cadherin expression in atrial fibrillation. J Am Coll Cardiol, 2010. 55(5): p. 469-80.11. Adam, O., K. Theobald, D. Lavall, et al., Increased lysyl oxidase expression and collagen cross-linking during atrial fibrillation. J Mol Cell Cardiol, 2011. 50(4): p. 678-85.12. Zhao, Y., K. Tang, X. Tianbao, J. Wang, J. Yang, and D. Li, Increased serum lysyl oxidase-like 2 levels correlate with the degree of left atrial fibrosis in patients with atrial fibrillation. Biosci Rep, 2017. 37(6).13. Di Tullio, M.R., M. Qian, J.L.P. Thompson, et al., Left atrial volume and cardiovascular outcomes in systolic heart failure: effect of antithrombotic treatment. ESC Heart Fail, 2018. 5(5): p. 800-808.14. Goette A, Kalman JM, Aguinaga L, et al. EHRA/HRS/APHRS/SOLAECE expert consensus on atrial cardiomyopathies: Definition, characterization, and clinical implication. Europace. 2016;18(10):1455-1490.15. Kottkamp H. Fibrotic atrial cardiomyopathy: a specific disease/syndrome supplying substrates for atrial fibrillation, atrial tachycardia, sinus node disease, AV node disease, and thromboembolic complications. J Cardiovasc Electrophysiol. 2012;23(7):797-799.16. Barry-Hamilton V, Spangler R, Marshall D, et al. Allosteric inhibition of lysyl oxidase-like-2 impedes the development of a pathologic microenvironment. Nat Med. 2010;16(9):1009-1017.17. Barker HE, Bird D, Lang G, Erler JT. Tumor-secreted LOXL2 activates fibroblasts through FAK signaling. Mol Cancer Res. 2013;11(11):1425-1436.18. Atsawasuwan P, Mochida Y, Parisuthiman D, Yamauchi M. Expression of lysyl oxidase isoforms in MC3T3-E1 osteoblastic cells. Biochem Biophys Res Commun. 2005;327(4):1042-1046.19. López B, González A, Hermida N, et al. Role of lysyl oxidase in myocardial fibrosis: from basic science to clinical aspects. Am J Physiol Heart Circ Physiol. 2010;299(1):H1-H9 .20. Chen Y, Guo H, Xu D, et al. Left atrial fibrosis in atrial fibrillation: mechanisms, clinical evaluation and management. J Cell Mol Med. 2021;25(6):2764-2775.21. Zhang Y, Zhang S, Li B, et al. LOXL2-mediated matrix remodeling in metastasis and mammary gland involution. Cancer Res. 2020;80(18):3875-3887.22. Wang X, Li Z, Liu H, et al. LOXL2 as a potential therapeutic target in atrial fibrillation. J Am Heart Assoc. 2021;10(12):e020123 . All Patients (N=197) Breakthrough Stroke p Value No (n=173) YES (n=24) Anthropometrics Age, yrs 67.3 ± 10.1 66.3 ± 9.9 74.5 ± 8.0 < 0.001 Male 113 (57.4%) 100 (57.8%) 13 (54.2%) 0.907 BMI, kg/m2 25.2 ± 3.9 25.3 ± 3.9 24.5 ± 4.0 0.417 Current smoker 74 (37.6%) 63 (36.4%) 11 (45.8%) 0.504 Regular alcohol consumption 42 (21.3%) 36 (20.8%) 6 (25.0%) 0.838 Rhythm in 0.615 Sinus 112 (56.9%) 100 (57.8%) 12 (50.0%) AF 85 (43.1%) 73 (42.2%) 12 (50.0%) Rhythm out 0.005 Sinus 189 (95.9%) 169 (97.7%) 20 (83.3%) AF 8 (4.1%) 4 (2.3%) 4 (16.7%) Comorbidities & Medical history Stoke history 20 (10.2%) 14 (8.1%) 6 (25.0%) 0.027 Hypertension 138 (70.1%) 121 (69.9%) 17 (70.8%) 1.000 Dyslipidemia 154 (78.2%) 131 (75.7%) 23 (95.8%) 0.049 Diabetes 55 (27.9%) 48 (27.7%) 7 (29.2%) 1.000 CAD 65 (33.0%) 54 (31.2%) 11 (45.8%) 0.232 Previous MI 18 (9.1%) 16 (9.2%) 2 (8.3%) 1.000 CABG 9 (4.6%) 6 (3.5%) 3 (12.5%) 0.143 PCI 30 (15.2%) 25 (14.5%) 5 (20.8%) 0.608 COPD 5 (2.5%) 3 (1.7%) 2 (8.3%) 0.217 Carotid artery disease 8 (4.1%) 3 (1.7%) 5 (20.8%) < 0.001 Peripheral vascular disease 10 (5.1%) 8 (4.6%) 2 (8.3%) 0.780 CKD 57 (28.9%) 45 (26.0%) 12 (50.0%) 0.029 Treatment Cather ablation 150 (76.1%) 137 (79.2%) 13 (54.2%) 0.015 Antiarrhythmic 92 (46.7%) 85 (49.1%) 7 (29.2%) 0.105 Anticoagulation 192 (97.5%) 169 (97.7%) 23 (95.8%) 1.000 Warfarin 87 (44.2%) 73 (42.2%) 14 (58.3%) 0.203 NOAC 105 (53.3%) 96 (55.5%) 9 (37.5%) 0.151 Antiplatelets 48 (24.4%) 35 (20.2%) 13 (54.2%) 0.001 Beta-blocker 93 (47.2%) 84 (48.6%) 9 (37.5%) 0.425 Statins 137 (69.5%) 58 (75.3%) 20 (69.0%) 0.184 ACE inhibitor/ARB 82 (41.6%) 73 (42.2%) 9 (37.5%) 0.829 CHA2DS2VASC score 0.005 Low Risk 18 (9.1%) 18 (10.4%) 0 (0.0%) Medium Risk 38 (19.3%) 36 (20.8%) 2 (8.3%) High Risk 141 (71.6%) 119 (68.8%) 22 (91.7%) Laboratory test & echocardiography LOXL2, pg/ml 296.0 ± 118.9 278.8 ± 81.4 419.8 ± 229.4 0.006 hsCRP, mg/L 2.7 ± 3.2 2.6 ± 3.0 3.5 ± 4.1 0.301 BNP, ng/ml 653.4 ± 1,328.3 566.2 ± 1,270.6 1,282.1 ± 1,578.8 0.043 TSH, mIU/L 2.5 ± 3.5 2.5 ± 3.7 2.7 ± 1.8 0.753 TnI, ng/ml 2.8 ± 10.5 3.1 ± 11.1 1.0 ± 2.7 0.036 BUN, mmol/L 363.1 ± 98.6 358.2 ± 95.4 397.7 ± 115.6 0.122 LA diameter, mm 39.6 ± 5.3 39.3 ± 4.9 41.9 ± 7.0 0.090 Values are n (%), or mean ± SD. Bold values indicate statistical significance. Diabetes was defined as plasma glucose >126 mg/dl. Dyslipidemia was defined by the presence of hypercholesterolemia (serum cholesterol >200 mg/dl) and/or hypertriglyceridemia (serum triglycerides >150 mg/dl). CKD was defined as eGFR <60 ml/min/1.73 m2. ACE inhibitors/ARB = angiotensin converting enzyme inhibitors/angiotensin receptor blockers; AF = atrial fibrillation; BMI = body mass index; BNP = B-type natriuretic peptide; BUN = blood urea nitrogen; CABG = coronary artery bypass graft; CAD = coronary artery disease; CKD = chronic kidney disease; COPD = chronic obstructive pulmonary disease; hsCRP = high-sensitivity C-reactive protein; LA diameter = left atrial diameter ; LOXL2 = lysyl oxidase like 2; NOAC = non-vitamin K antagonist oral anticoagulant; PCI = percutaneous coronary intervention; TnI = peak troponin I; TSH = thyroid-stimulating hormone. Variable Univariate Multivariate HR (95% CI) P Value HR (95% CI) p Value Age 1.120 (1.055-1.188) < 0.001 1.131 (2.795-3.474) 0.012 Rhythm 5.854 (1.997-17.161) 0.001 Stoke history 3.543 (1.406-8.928) 0.007 Carotid artery disease 7.874 (2.925-21.196) < 0.001 CKD 2.596 1.166-5.779) 0.019 Cather ablation 0.339 (0.152-0.758) 0.008 Antiplatelets 4.106 (1.834-9.169) < 0.001 3.206 (2.923-14531) 0.037 CHA2DS2VASC 1.382 (1.119-1.707) 0.003 LOXL2 1.006 (1.004-1.008) < 0.001 1.008 (2.730-2.751) < 0.001 BNP, ng/ml 1.0002 (1.0000-1.0003) < 0.001 BUN, mmol/L 1.004 (1.000-1.007) 0.047 LA diameter, mm 1.082 (1.010-1.158) 0.024 Dyslipidemia 6.807 (0.919-50.410) 0.060 CABG 3.223 (0.961-10.813) 0.058 COPD 3.959 (0.929-16.865) 0.063 Antiarrhythmic 0.440 (0.182-1.060) 0.067 Bold values indicate statistical significance after multiple testing correction. Abbreviations as in Table 1. LOXL2- (n=99) LOXL2+ (n =98） p Value Stroke 7 (7.1%) 17 (17.3%) 0.047 Anthropometrics Age, yrs 67.4 ± 10.8 67.3 ± 9.3 0.946 Male 63 (63.6%) 50 (51.0%) 0.100 BMI, kg/m2 24.8 ± 3.9 25.6 ± 4.0 0.157 Current smoker 37 (37.4%) 37 (37.8%) 1.000 Regular alcohol consumption 22 (22.2%) 20 (20.4%) 0.891 Type of AF 0.051 Paroxysmal 49 (49.5%) 63 (64.3%) Persistent 50 (50.5%) 35 (35.7%) Rhythm 0.272 Sinus 97 (98.0%) 92 (93.9%) AF 2 (2.0%) 6 (6.1%) Comorbidities & Medical history Stoke history 6 (6.1%) 14 (14.3%) 0.094 Hypertension 66 (66.7%) 72 (73.5%) 0.375 Dyslipidemia 74 (74.7%) 80 (81.6%) 0.319 Diabetes 24 (24.2%) 31 (31.6%) 0.319 CAD 33 (33.3%) 32 (32.7%) 1.000 Previous MI 7 (7.1%) 11 (11.2%) 0.445 CABG 7 (7.1%) 2 (2.0%) 0.177 PCI 15 (15.2%) 15 (15.3%) 1.000 COPD 3 (3.0%) 2 (2.0%) 1.000 Carotid artery disease 5 (5.1%) 3 (3.1%) 0.729 Peripheral vascular disease 5 (5.1%) 5 (5.1%) 1.000 CKD 26 (26.3%) 31 (31.6%) 0.500 Treatment Cather ablation 78 (78.8%) 72 (73.5%) 0.479 Antiarrhythmic 44 (44.4%) 48 (49.0%) 0.621 Anticoagulation 95 (96.0%) 97 (99.0%) 0.371 Warfarin 44 (44.4%) 43 (43.9%) 1.000 NOAC 51 (51.5%) 54 (55.1%) 0.718 Antiplatelets 25 (25.3%) 23 (23.5%) 0.900 Beta-blocker 43 (43.4%) 50 (51.0%) 0.356 Statins 69 (69.7%) 68 (69.4%) 1.000 ACE inhibitor/ARB 40 (40.4%) 42 (42.9%) 0.838 CHA2DS2VASC score 0.087 Low Risk 11 (11.1%) 7 (7.2%) Medium Risk 22 (22.2%) 16 (16.3%) High Risk 66 (66.7%) 75 (76.5%) Laboratory test & echocardiography hsCRP, mg/L 2.5 ± 2.9 2.8 ± 3.5 0.531 BNP, ng/ml 671.6 ± 1,580.4 635.1 ± 1,020.4 0.848 TSH, mIU/L 2.2 ± 1.9 2.9 ± 4.6 0.139 TnI, ng/ml 2.7 ± 4.6 3.0 ± 14.1 0.832 BUN, mmol/L 358.0 ± 101.6 368.2 ± 95.6 0.471 LA diameter, mm 39.1 ± 4.7 40.1 ± 5.7 0.150 Values are n (%), or mean ± SD. Bold values indicate statistical significance. Abbreviations as in Table 1. Statistic Parameter Hazard Ratio 95% CI p Value Unadjusted Cox regression LOXL2-(reference) 1.00 LOXL2+ 12.56 2.89 - 484.85 0.036 Adjusted Cox regression LOXL2-(reference) 1.00 LOXL2+ 1.98 3.33 - 26.27 0.007 Added predictive value NRI 0.15 0.10-0.20 <0.001 *Hazard ratios were adjusted by the Model A: CHA2DS2VASc score. Patients were classified as LOXL2+ or LOXL2-, respectively, as defined in the main text. NRI = net reclassification improvement. Figure Legend Figure 1. Baseline Characteristics and Serum LOXL2 Levels in anticoagulated PAF Patients With and Without Breakthrough Stroke. (A) Distribution of CHA2DS2-VASc scores in the study cohort; (B) Anticoagulation status of the study cohort; (C) Comparison of serum LOXL2 levels between stroke and non-stroke groups; (D) Correlation analysis between serum LOXL2 levels and BNP, TNI, and CHA2DS2VASc score. BNP = B-type natriuretic peptide; LOXL2 = lysyl oxidase like 2; PAF = paroxysmal atrial fibrillation; TNI = peak troponin I. Figure 2. LOXL2 as an Independent Risk Factor for Breakthrough Stroke in Patients with PAF. (A) Forest plot of multivariate Cox regression analysis showing independent predictors of breakthrough stroke in PAF patients; (B) breakthrough stroke incidence in patients stratified by LOXL2 levels (cutoff: 275.9 pg/mL); (C) Kaplan-Meier survival analysis comparing stroke-free survival between LOXL2+ and LOXL2− groups; (D) Nelson-Aalen cumulative hazard curves demonstrating an increased cumulative hazard for breakthrough stroke in the LOXL2+ group, with a cumulative hazard ratio of 2.56. AAD: antiarrhythmic drugs; BUN = blood urea nitrogen; CABG = coronary artery bypass grafting; CI = confidence interval; CKD = chronic kidney disease; COPD = chronic obstructive pulmonary disease; HR = hazard ratio; LAD: left atrial diameter; RFCA = radiofrequency catheter ablation. other abbreviations as in Figure 1 . Figure 3. LOXL2 Enhances Breakthrough Stroke Risk Prediction in Patients with PAF. (A) ROC curves comparing the predictive performance of four models: Model A (CHA2DS2VASc score alone), Model B (LOXL2 alone), Model C (LOXL2 + age), and Model D (LOXL2 + CHA2DS2VASc score); (B) Dynamic AUC analysis assessing the temporal stability of the predictive models; (C) DCA comparing the clinical utility of the models; (D) NRI analysis demonstrated that Model C significantly improved breakthrough stroke risk stratification compared to Model A. AUC = area under the curve; DCA = Decision curve analysis; NRI = net reclassification improvement. ROC = Receiver operating characteristic. other abbreviations as in Figure 1 . Supplementary Material File (image1.emf) Download 1.24 MB File (image2.emf) Download 1.49 MB File (image3.emf) Download 1.24 MB Information & Authors Information Version history V1 Version 1 04 March 2025 Copyright This work is licensed under a Non Exclusive No Reuse License. Keywords basic: atrial fibrillation/atrial arrhythmias clinical: catheter ablation – atrial fibrillation Authors Affiliations Fang Liu 0000-0002-4528-6087 Beijing Anzhen Hospital Affiliated to Capital Medical University View all articles by this author Zhitian Li Beijing Shijitan Hospital Capital Medical University View all articles by this author Ruolin Wang Beijing Anzhen Hospital Affiliated to Capital Medical University View all articles by this author Shanshan Meng Peking University Third Hospital View all articles by this author Ziting Wu Peking University Institute for Global Health and Development View all articles by this author Dongtao Zhou Beijing Anzhen Hospital Affiliated to Capital Medical University View all articles by this author Qianran Luan The Third Affiliated Hospital of Chengdu University of Chinese Medicine View all articles by this author Ying Dong Beijing Chaoyang Hospital Affiliated to Capital Medical University View all articles by this author Chen-Xi Jiang Beijing Anzhen Hospital Affiliated to Capital Medical University View all articles by this author Ribo Tang Beijing Anzhen Hospital Affiliated to Capital Medical University View all articles by this author Wei Wang Beijing Anzhen Hospital Affiliated to Capital Medical University View all articles by this author Xin Zhao 0000-0003-3663-6097 Beijing Anzhen Hospital Affiliated to Capital Medical University View all articles by this author Changyi Li Beijing Anzhen Hospital Affiliated to Capital Medical University View all articles by this author Tong Liu 0000-0001-5372-7236 Beijing Anzhen Hospital Affiliated to Capital Medical University View all articles by this author Yue-Xin Jiang Beijing Anzhen Hospital Affiliated to Capital Medical University View all articles by this author Mengmeng Li Beijing Anzhen Hospital Affiliated to Capital Medical University View all articles by this author Deyong Long 0000-0003-4604-5346 Beijing Anzhen Hospital Affiliated to Capital Medical University View all articles by this author Yuanfeng Gao [email protected] Beijing Chaoyang Hospital Affiliated to Capital Medical University View all articles by this author Metrics & Citations Metrics Article Usage 219 views 180 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Fang Liu, Zhitian Li, Ruolin Wang, et al. LOXL2 as a Predictive Biomarker for Breakthrough Stroke Risk in Paroxysmal Atrial Fibrillation Patients. Authorea . 04 March 2025. DOI: https://doi.org/10.22541/au.174110600.06008991/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.174110600.06008991/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9ff4148bccb806f7',t:'MTc3OTM3MTc0Mg=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();