# Visceral Embolic Events in Atrial Fibrillation: A Systematic Review and Meta-Analysis of Incidence, Mortality, and Risk Prediction

## TL;DR

This study reviews and analyzes the occurrence and risks of visceral embolic events in patients with atrial fibrillation, highlighting their under-recognized impact and potential risk factors.

## Contribution

The study provides the first reliable pooled estimate of non-AMI visceral embolic events in atrial fibrillation patients using event-definition-based subgroup analyses.

## Key findings

- Visceral embolic events occur in approximately 1–6% of atrial fibrillation patients, with a pooled prevalence of 1.6% for non-AMI events.
- Mortality from visceral embolic events varies widely, from 17% to 64%, depending on patient age and concurrent acute myocardial infarction.
- Left atrial enlargement and elevated D-dimer are potential predictors of visceral embolic events, though further validation is needed.

## Abstract

Background: Visceral embolic events (VEE), including mesenteric, splenic, and renal infarctions, represent understudied complications of atrial fibrillation (AF) often subsumed within broader systemic embolic event categories. The 2024 European Society of Cardiology guidelines introduced the CHA2DS2-VA score, removing female sex as a risk modifier, with potential implications for non-cerebral embolic risk stratification. We systematically synthesized evidence on VEE incidence, mortality, and risk predictors in AF patients. Methods: We searched PubMed, Embase, Scopus, and Web of Science through September 2024 for studies reporting VEE in AF populations. Study quality was assessed using the Newcastle–Ottawa Scale (NOS). Due to substantial heterogeneity when pooling all prevalence studies (I2 = 99.6%), we performed event-definition-based subgroup analyses. Random-effects meta-analyses were conducted using DerSimonian–Laird methods with 95% prediction intervals. Sensitivity analyses excluded studies with NOS scores < 7 to assess robustness. Results: We identified 12 studies including 329,128 patients. Quality assessment revealed a mean NOS score of 6.7 ± 1.6 (range: 4–9), with 75% of studies achieving good-to-excellent ratings. For non-AMI visceral embolic events (splenic, renal, mesenteric infarctions), subgroup meta-analysis of three studies (n = 548) yielded a pooled prevalence of 1.6% (95% CI: 0.0–3.2%, I2 = 45.4%, p = 0.160), representing a 54.2 percentage point reduction in heterogeneity compared to pooling all event types. Sensitivity analysis excluding moderate-quality studies confirmed robust findings (pooled prevalence 2.7%, 95% CI: 0.0–6.8%, I2 = 70.3%). Incidence rates ranged from 0.36 to 3.48 per 1000 person-years across three cohort studies (I2 = 99.4%), reflecting temporal and geographic variation. Mortality varied substantially by patient population: 64.0% in-hospital mortality among elderly patients with concurrent acute myocardial infarction (AMI) versus 17.4% in younger cohorts with isolated non-AMI VEE. Potential predictors included left atrial enlargement (OR range: 2.1–4.3), elevated D-dimer (OR: 3.2), and higher CHA2DS2-VASc scores (OR: 1.3 per point increase), though validation in independent cohorts is lacking. Conclusions: Visceral embolic events occur in approximately 1–6% of AF patients, with mean prevalence of 1.6% for non-AMI events based on moderate-heterogeneity meta-analysis. Event-definition-based subgrouping successfully reduced heterogeneity from 99.6% to 45.4%, providing the first reliable pooled estimate for this outcome. Mortality ranges widely (17–64%) depending on concurrent AMI and patient age. Potential predictors including left atrial enlargement and elevated D-dimer require prospective validation before clinical implementation. These findings suggest VEE may warrant enhanced clinical awareness and individualized risk assessment strategies in AF management, pending validation in prospective studies.