Stent-assisted coiling improves aneurysm occlusion but raises ischemic risk in meta-analysis

Stent-assisted coiling of unruptured brain aneurysms delivers better long-term vessel sealing than conventional coiling, but it nearly doubles the risk of ischemic complications, according to a systematic review and meta-analysis published in the Egyptian Journal of Neurosurgery. The analysis pooled eight studies covering 7,682 patients and found that while stents pushed complete aneurysm occlusion at follow-up significantly higher, they offered no clear advantage in recurrence, procedural rupture, or bleeding, leaving clinicians with a sharp trade-off between angiographic durability and clot-related harm.

The research team, led by Anfal Binte Amin and Abdullah Afridi, registered the protocol in PROSPERO and followed PRISMA reporting guidelines. They searched PubMed, Embase, and the Cochrane Library from database inception through July 30, 2025, capturing randomized controlled trials and observational studies that compared stent-assisted coiling (SAC) with non-stent-assisted coiling (non-SAC) in adults with unruptured intracranial aneurysms. Two reviewers independently screened titles and abstracts, extracted data into a standardized spreadsheet, and resolved disagreements with a third reviewer. From 1,427 initial records, 65 full-text reports were assessed and eight studies survived: one randomized controlled trial and seven observational cohorts conducted in Canada, France, the United States, Italy, South Korea, and Japan, with follow-up durations spanning one to ten years.

The patient population was unevenly split. Of the 7,682 participants, 2,064 underwent SAC and 5,618 received non-SAC, and mean ages across the studies ranged from 54 to 64 years. The reviewers assessed observational quality with the Newcastle–Ottawa Scale, where scores between 5 and 8 indicated moderate-to-high quality, and applied the revised Cochrane Risk of Bias 2 tool to the single randomized trial. That trial was judged to carry an overall high risk of bias, driven by concerns about deviations from intended interventions and a high risk of bias in outcome measurement, a limitation that shadowed the entire synthesis.

Statistically, the team used Review Manager 5.4 with Mantel-Haenszel random-effects models, pooling risk ratios with 95 percent confidence intervals and flagging heterogeneity with the I² statistic. On the headline efficacy outcome, seven studies reported complete occlusion after stent assistance, and the pooled estimate favored SAC: a risk ratio of 1.19 with a confidence interval of 1.09 to 1.29 and a P value below 0.0001, with no observed heterogeneity. In plain terms, patients whose aneurysms were secured with a stent were roughly 19 percent more likely to show complete occlusion at follow-up angiography than those coiled without one.

The initial angiographic picture was murkier. Complete occlusion before stent assistance, pooled across all eight studies, showed no statistically significant difference between the groups (risk ratio 1.16, 95 percent confidence interval 0.94 to 1.42, P = 0.17), and the heterogeneity was substantial at I² = 87 percent. The authors suggested that differences in aneurysm size, neck width, morphology, stent generation, coil strategy, operator experience, imaging protocols, and treatment era could explain that dispersion. The divergence between the two occlusion time points hints at a time-dependent effect: stents may scaffold the aneurysm neck, permit denser coil packing, alter intra-aneurysmal flow, and promote progressive thrombosis and endothelialization, so the angiographic advantage accrues over months rather than appearing immediately.

Safety told a different story. Ischemic events, reported in seven studies, occurred significantly more often with SAC, with a risk ratio of 1.81 (95 percent confidence interval 1.51 to 2.19, P < 0.00001, I² = 0 percent), an 81 percent relative increase. The discussion notes that without pooled absolute risks, the analysis could not calculate risk differences, numbers needed to treat, or numbers needed to harm, nor could it distinguish transient symptoms from imaging-only infarcts or disabling strokes, or separate periprocedural from delayed events. Hemorrhagic events, pooled across six studies, showed no significant difference (risk ratio 1.06, 95 percent confidence interval 0.62 to 1.78, P = 0.84), and intraprocedural rupture, drawn from only three cohorts, was similarly inconclusive (risk ratio 0.49, 95 percent confidence interval 0.09 to 2.75, P = 0.41), a wide interval that permits either substantial benefit or substantial harm. Recurrence, reported in seven studies, also showed no significant difference (risk ratio 0.87, 95 percent confidence interval 0.64 to 1.18, P = 0.36).

The ischemic signal is the central clinical trade-off. A permanent intraluminal stent adds a thrombogenic surface and requires dual antiplatelet therapy, and the authors point out that antiplatelet selection, responsiveness, duration, and adherence may modify both ischemic and bleeding risks. Inadequate platelet inhibition may permit device thrombosis, while prolonged dual therapy may raise bleeding risk. Confounding by indication is another live concern: SAC may have been preferentially deployed for anatomically complex, wide-neck aneurysms that carry greater technical risk with conventional coiling, since coil instability in wide-neck lesions can cause wire prolapse into the parent vessel, incomplete occlusion, and higher recurrence. Moderate-to-high Newcastle–Ottawa scores do not eliminate that confounding, and the unequal group sizes, 2,070 versus 5,631 patients in the strength-and-limitations accounting, leave the synthesis vulnerable to selection bias and residual confounding.

The authors are explicit that statistical significance in a radiographic endpoint does not equal patient benefit. The follow-up occlusion advantage was not accompanied by lower recurrence and was offset by the ischemic increase. Functional status, disability, mortality, quality of life, post-treatment rupture, and retreatment were unavailable in the pooled data, so the evidence cannot determine whether the radiographic improvement translates into a net clinical benefit. They also note that mixing patient-level and aneurysm-level units without clustering adjustment may overstate precision, that no subgroup, sensitivity, meta-regression, publication-bias, or certainty-of-evidence analyses were reported, and that eight studies provide limited power to detect publication bias. The single randomized trial’s high risk of bias, particularly in outcome measurement, further tempers confidence.

Clinically, the authors argue against routine SAC solely to improve follow-up occlusion. They align with European Stroke Organisation guidance that treatment is warranted only when estimated rupture risk exceeds procedural risk within a shared decision-making framework. SAC may remain reasonable when neck geometry or coil stability makes non-stent coiling unsuitable, but the review conducted no subgroup analyses and cannot identify a favorable subgroup. Alternative approaches, including simple or balloon-assisted coiling, dual-microcatheter techniques, intrasaccular devices such as WEB for selected wide-neck bifurcation aneurysms, flow diversion, surgical clipping, and surveillance, remain anatomy- and context-dependent options rather than direct comparators in this meta-analysis. Treatment selection, the authors write, should weigh age, comorbidity, antiplatelet tolerance and adherence, planned surgery, operator expertise, follow-up capacity, patient preference, device and medication costs, and feasibility in resource-limited settings.

Looking forward, the team calls for multicenter randomized trials with anatomical eligibility criteria, standardized protocols, and blinded core-laboratory assessment. Such trials should prespecify initial and long-term angiographic time points, standardize definitions of recurrence, retreatment, symptomatic ischemia, imaging-only lesions, and hemorrhage, and prioritize patient-important endpoints including disabling stroke, modified Rankin Scale scores, mortality, quality of life, post-treatment rupture, retreatment, and cost-effectiveness, with absolute risks and time-to-event outcomes reported over three to five years. Prespecified interaction analyses by anatomy, stent type, antiplatelet regimen, treatment era, and center experience, ideally through an individual-patient-data meta-analysis, would help identify which patients truly benefit. Until such evidence arrives, the current data support selective, individualized use of stent-assisted coiling rather than its routine adoption for every unruptured intracranial aneurysm.

Subject of Research: Systematic review and meta-analysis of stent-assisted versus non-stent-assisted coiling for unruptured intracranial aneurysms

Article Title: Stent assisted coiling versus non-stent assisted coiling for unruptured intracranial aneurysms: a systematic review and meta-analysis for clinical and angiographic outcomes

Article References: Amin, A. B., Yousaf, F., Hussain, S. M. A., Zafar, H. B., Huzaifah Khattak, M., Sajjad, F., Moiz, A., Moez, A., Khattak, A., Ali, B. M., Mahparah, F. N. U., Shah, L., Ali, M., Nauman, M. F., Ali, A., Abed, A., Ullah, M., Mehmood Jan, M., Zayed, M., … Afridi, A. (2026). Stent assisted coiling versus non-stent assisted coiling for unruptured intracranial aneurysms: a systematic review and meta-analysis for clinical and angiographic outcomes. Egyptian Journal of Neurosurgery, 41(1), Article 176. https://doi.org/10.1186/s41984-026-00684-x

Image Credits: AI Generated

DOI: 10.1186/s41984-026-00684-x

Keywords: unruptured intracranial aneurysms, stent-assisted coiling, non-stent-assisted coiling, meta-analysis, complete occlusion, ischemic events, intraprocedural rupture, recurrence, hemorrhagic events, dual antiplatelet therapy, observational studies, endovascular treatment