Reported Complication Profiles of Liquid Embolic Agents for Peripheral Arteriovenous Malformation Embolization: A Pooled Analysis of Single-Arm Studies

Abstract The aim of the study is to systematically summarize the reported complication profiles of ethanol, ethylene vinyl alcohol copolymer (EVOH), and n-butyl cyanoacrylate (nBCA) used for embolization of peripheral arteriovenous malformations (AVM), and to descriptively characterize their safety. A systematic search of PubMed, ClinicalKey, Web of Science, and Google Scholar identified original studies reporting complications of peripheral AVM embolization. Eligible studies enrolled at least three patients and reported outcomes for a single liquid agent or provided agent-specific data. Complications were standardized to the modified Cardiovascular and Interventional Radiological Society of Europe classification and pooled for each agent separately using single-arm random-effects models, with rates summarized as pooled proportions, 95% confidence intervals, and prediction intervals. Because all included studies were single-arm and non-comparative, between-agent and between-site contrasts (Q test) were treated as exploratory and hypothesis-generating rather than as formal head-to-head comparisons. Forty-one single-arm studies were included (ethanol, n = 16, 494 patients; EVOH, n = 17, 274 patients; nBCA, n = 8, 86 patients). Pooled major complication rates were 5.65% (95% CI: 3.03–9.02; I 2 = 50.7%) for ethanol, 3.99% (2.02–6.58; I 2 = 0%) for EVOH, and 1.64% (0.06–5.28; I 2 = 0%) for nBCA, with overlapping confidence intervals and no difference on exploratory testing (p = 0.20). Pooled minor complication rates were 47.00% (34.44–59.75; I 2 = 87.7%) for ethanol, 15.69% (8.87–24.00; I 2 = 64.2%) for EVOH, and 23.01% (6.21–46.33; I 2 = 74.5%) for nBCA. The highest pooled minor complication rate was observed in the ethanol series; however, heterogeneity was highly considerable. Across non-comparative case series, all three agents were associated with generally low pooled major complication rates, while ethanol series reported the highest pooled minor complication rate. These pooled estimates are descriptive and are confounded by differences in lesion site, angioarchitecture, treatment intent, and reporting; heterogeneity further limits inference. Prospective, comparative studies using standardized outcome reporting are needed.