Next-generation Onyx DES for elective intracranial atherosclerosis: A meta-analysis

Abstract

Background

Intracranial atherosclerotic disease (ICAD) is a growing cause of ischemic stroke globally, with a disproportionately high burden in Asian, Black, and Hispanic populations. Despite advances in medical therapy, ICAD remains associated with high rates of recurrent stroke, prompting interest in durable endovascular solutions. This study aims to systematically evaluate the current evidence on the safety and efficacy of elective intracranial stenting in adult patients with symptomatic ICAD using Onyx drug-eluting balloon-mounted stents (Onyx DES).

Methods

A meta-analysis was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Five studies were initially identified, with one excluded due to overlapping cohorts, resulting in four final studies included for analysis. Meta-analysis was conducted utilizing random-effects models for pooled event rates (95% CIs) and weighted means.

Results

Of 153 articles initially identified, four high-quality studies encompassing 314 patients met inclusion criteria. The mean (±SD) age was 64.6 (±12.4) years, with a predominance of males (70.7%) and common vascular comorbidities such as hypertension (85.7%) and diabetes (60.0%). Lesions were nearly equally distributed between the anterior and posterior circulations. Periprocedural complications were infrequent (1.0%), including one hemorrhagic stroke and one fatal aneurysm rupture. The 30-day complication rate remained low at 5%, involving strokes, deaths, and TIAs. However, follow-up at six months and beyond revealed rising rates of strokes, TIAs, and in-stent restenosis, reaching 9% at six months and persisting through one year.

Conclusion

This descriptive meta-analysis suggests that Onyx DES may offer promising results in the treatment of symptomatic ICAD, with lower early complication rates reported. However, larger prospective studies are needed to confirm these observations and evaluate long-term efficacy and safety.

Introduction

Intracranial atherosclerotic disease (ICAD) is a major contributor to strokes across the globe, with an estimated incidence of 20 to 40 cases of ICAD-related cerebral infarctions per 100,000 people worldwide.^1,2 With the development of endovascular options, improved medical treatment, and advanced imaging, the management of ICAD has been changing. Over the previous six years, the use of endovascular therapy for symptomatic ICAD has significantly increased in the United States. ³ The unsatisfactory findings of two pivotal trials, SAMMPRIS and VISSIT, which revealed greater rates of periprocedural complications in the stenting arms when compared to medical therapy alone, dampened the initial optimism for percutaneous transluminal angioplasty and stenting (PTAS).^4,5

However, later studies like the WEAVE trial, which involved the use of the Wingspan (Stryker) stent, revived interest in stenting as a treatment option due to its lower rate of periprocedural stroke or death (2.6%). ⁶

Intracranial stenting can be performed using a variety of stent types, including drug-eluting balloon-mounted stents (DES)^{7,8,9,10–11} and self-expandable stents (SES).^6,12 The latter have a higher rate of in-stent restenosis (ISR) and a lower radial force, making them less effective at achieving the desired luminal dilatation, particularly in patients with calcified lesions.^13,14

The Onyx DES (Onyx Resolute and Onyx Frontier) was initially developed to treat patients with coronary artery disease. It helps prevent ISR by delivering a high dose of the anti-proliferative agent Zotarolimus directly to the vascular wall.^15,16 Recently, the Onyx DES stent has gained interest for use in the endovascular treatment of ICAD. Preliminary observational studies and multicentre retrospective analyses have suggested that Onyx DES offers high technical success rates with lower periprocedural complication rates and reduced ISR compared to older technologies.^{17,18,19,20–21} Compared to other balloon-mounted stents, Onyx DES offers enhanced navigability in tortuous intracranial anatomy and has demonstrated lower early complication rates in emerging clinical data. While off-label for intracranial use, Onyx DES is among the more extensively utilized and studied DES platforms in current neurointerventional practice. ²¹ Despite promising early data, the long-term efficacy and safety of Onyx DES in symptomatic ICAD (sICAD) remain inadequately characterized.

The aim of this meta-analysis is to provide a descriptive evaluation of the safety and effectiveness of the Onyx DES in the treatment of patients with elective sICAD. Specifically, we focus on procedural success rates, the incidence of ISR, and both short-term and long-term risks of stroke or death following stent placement. By synthesizing available data, we hope to better understand the evolving role of balloon-mounted DES in this high-risk patient population, many of whom have few remaining options after failing medical therapy. Ultimately, this descriptive analysis may help inform future research and guide more personalized decision-making in the treatment of sICAD.

Methods

Search

The study followed the Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) 2020 guidelines. ²² On 30 March 2025, we conducted searches on PubMed, Scopus, Web of Science, Cochrane, and ClinicalTrials.gov from January 2015 to March 2025 with the main search strategy: (“Onyx” OR “Resolute” OR “Onyx Resolute” OR “Resolute Onyx” OR “Onyx Frontier” OR “Frontier Stent” OR “Zotarolimus-eluting Stent” OR “Drug-eluting Stent” OR “Frontier”) AND (“Intracranial Atherosclerosis” OR “Intracranial Atherosclerotic Disease” OR “ICAD” OR “Intracranial Stenosis”) AND (“Intracranial Stenting” OR “Endovascular Therapy” OR “Angioplasty” OR “Revascularization”). Please refer to the PRISMA flowchart for the specific results from the searches (Figure 1).

Figure 1.

PRISMA flow chart for study selection.

Inclusion criteria and exclusion criteria

We included studies involving adults (≥18 years) with sICAD who underwent elective stenting with Onyx DES (Onyx Resolute and Onyx Frontier). Studies published between 2015 and 2025 were eligible, with no restrictions on sex, race, ethnicity, language, or country. We excluded case reports (<5 patients), conference abstracts without full-text availability, reviews, pediatric populations (<18 years), stenting for non-atherosclerotic etiologies, studies with rescue Onyx stenting, and studies evaluating non-Onyx stents. Studies with rescue Onyx stenting were excluded to avoid bias from emergent, non-elective cases, which differ in risk profile and outcomes. Non-Onyx stent studies were excluded to maintain device-specific analysis and ensure consistency in evaluating the safety and efficacy of Onyx stents.

Data extraction

Two investigators (AB and TAH) independently screened the titles and abstracts of the articles to exclude irrelevant studies. Then, full-text screening was conducted independently by the same investigators to determine eligible articles to be included in the systematic review and meta-analysis. All disagreements were resolved by our supervisor (SP). Data extracted included: author-name, year published, study methodology, sample size, mean age, sex, ethnicity, comorbidities data including hypertension (HTN), hyperlipidemia (HLD), diabetes mellitus (DM), smoking status, and time from symptoms onset to intracranial stenting (days). Outcomes extracted included periprocedural complications and postoperative complications within 30 days, as well as the occurrence of ISR (≥50% stenosis), strokes, and death within 6 months and 12 months.

Outcomes of interest

Time from presenting symptoms to stenting was extracted from all studies. We considered complications that occurred during the intervention or within 72 h after stenting as periprocedural complications. Beyond the immediate periprocedural period, complications were tracked at multiple intervals. We grouped the complications into 3 time points. Firstly, the complications recorded from the 3rd to the 30th postoperative days were collected at the 30-day complications rate. Secondly, complications within a 6-month time frame. Lastly, complications were reported during the 1-year period.

Quality assessment

Studies were scored by two independent reviewers (AB and TAH) using the Newcastle-Ottawa Scale (NOS)²³ modified to suit the nature of single-arm stent trials, with a scale of 0–7 stars. Firstly, we replaced comparability domains with a Technical Quality criterion (1 star) to evaluate protocol standardization (e.g., antiplatelet consistency). Secondly, we prioritized angiographic outcomes (awarding two stars for core-lab adjudicated 1-year imaging vs. one star for clinical-only follow-up). Thirdly, we simplified selection bias assessment (two stars for multicenter studies; one star for single-center studies). High-quality studies (6–7 stars) featured multicenter designs, core-lab adjudication, long-term and angiographic follow-up, while moderate scores (4–5 stars) reflected single-center, without core-lab adjunctions and with shorter follow-up periods (Supplementary Table 7). Meta-analysis was performed on studies that exhibited high or moderate scores.

Statistical analysis

The Comprehensive Meta-Analysis Software version 3.0 was used for the meta-analysis. Results were presented as pooled event rates with a 95% CI calculated for dichotomous outcomes, and pooled weighted means for continuous variables. A random-effects model was selected for all meta-analyses to consider any clinical heterogeneity across studies, like procedural techniques and operator experience. These results were displayed using forest plots (Supplementary Figures S1–S4, Figures 2 and 3). Heterogeneity was assessed via I² statistics; values ≤50% were considered low heterogeneity, and ≥75% were considered high.²⁴ Sensitivity analysis was run for robustness, and no outliers were detected. Publication bias was assessed using the same software and depicted as Funnel plots with Egger's test, showing no significant publication bias for the time from symptoms onset to intracranial stenting (p = 0.29); periprocedural complications (p = 0.56), 30-day complications (p = 0.24), or 6-month complications (p = 0.47) (Supplementary Figures S5–S8, Supplementary Table S6). For studies reporting medians and IQRs, we used the Xiang et al.’s²⁵ method to convert to means and standard deviations.

Figure 2.

Forest plot of mean time from symptoms onset to intracranial stenting.

Figure 3.

The line chart illustrates the prevalence of post-procedural adverse events over a 12-month follow-up period, stratified by event type: total deaths, total stroke (ischemic/hemorrhagic), total TIA, and in-stent restenosis.

Results

The initial database search strategy yielded a total of 153 records, as follows: PubMed (n = 18), Scopus (n = 29), Web of Science (n = 24), ClinicalTrial.gov (n = 64), and Cochrane (n = 18), as shown in Figure 1. Firstly, 48 duplicates were removed, then 93 articles were excluded during title and abstract screening (Figure 1). Lastly, full-text screening yielded the exclusion of eight articles. A total of five studies^17–21 met the inclusion criteria for elective stenting with Onyx DES in sICAD patients aged 18 years or older. Through author correspondence, it was verified that the Hassan et al.’s study ¹⁸ extended the follow-up of the 2020 cohort, ¹⁷ whereas the Hassan et al.’s study ¹⁹ enrolled a separate set of symptomatic patients, allowing for appropriate data integration without overlap (Figure 1). As a result, Hassan et al.’s study ¹⁷ was excluded, and a total of four studies, reporting on 314 patients, were included in the final analysis (Figure 1).

Baseline characteristics were reported for 100% (N = 314/314) of patients. The pooled mean (±SD) age was 64.6 (±12.4) years (95% CI: 63.3–65.9, I² = 0%), 70.7% (N = 222/314) were men, and 38.9% (N = 122/314) were White, 33.1% (N = 104/314) were Hispanic and 6.1% (N = 19/314) were African (Table 1). Pooled prevalence of hypertension (HTN) was 85.7% (N = 269/314), hyperlipidemia (HLD) 70.1% (N = 220/314), diabetes mellitus (DM) 60.0% (N = 188/314), and smoking 32.0% (N = 100/314) (Table 1). A previous history of prior strokes or TIAs was reported in two studies with a total of 236 patients, a pooled prevalence of 64% (N = 151/236) for recurrent strokes and 35.2% (N = 83/236) for recurrent TIAs^18,21 (Table 1).

Table 1.

Baseline characteristics and comorbidities of included studies.

Variables	Siddiq/USA/2023 21	Hassan/USA/2023 18	Ravi/USA/ 2024 20	Hassan/USA/2024 19
	N = 132 (42.0%)	N = 58 (18.5%)	N = 104 (33.1%)	N = 20 (6.7%)
Age, mean ± SD	64.4 ± 12.2	63.7 ± 13.1	64.9 ± 10.9	66.8 ± 14
Gender (N, %)
Male	92 (69.7)	37 (63.8)	78 (74.5)	15 (75.0)
Female	40 (30.3)	21 (36.2)	26 (25.5)	5 (25.0)
Ethnicity (N, %)
White	58 (43.9)	4 (6.9)	56 (54.9)	4 (20.0)
Hispanic	–	54 (93.1)	34 (29.4)	16 (80.0)
African	11 (8.3)	–	7 (7.8)	–
Asian	–	–	1 (1)	–
Others	63 (47.7)	–	–	–
Comorbidities (N, %)
Hypertension	108 (81.8)	55 (94.8)	86 (84.3)	20 (100)
Diabetes	72 (54.6)	44 (75.9)	55 (49.1)	17 (85.0)
Smoking	44 (33.3)	6 (10.3)	49 (45.1)	4 (20.0)
Hyperlipidemia	89 (67.4)	39 (67.2)	73 (71.6)	19 (95.0)
Prior stroke	85 (64.4)	–	66 (64.7)	–
Prior TIA	47 (35.6)	–	35 (36.3)	–

Baseline vascular lesion distribution was reported across 3 studies^19–21 with a total of 256 patients (Table 2). One patient was presented with two distinct anterior circulation lesions ¹⁹ (total lesions = 257). Out of the 257 (100%) lesions, a total of 122 (47.5%) lesions were in the anterior circulation, 127 (49.4%) lesions were in posterior circulation, and symptomatic unidentified arteries were reported in 7 patients (2.7%)²⁰ (Table 2). The anatomical lesions were located in the vertebral artery (86 lesions, 33.5%), internal carotid artery (48 lesions, 18.7%), middle cerebral artery (44 lesions, 17.1%), basilar artery (40 lesions, 15.6%), and posterior cerebral artery (1 lesion, 0.4%) (Table 2).

Table 2.

Distribution of artery involvement in ICAD stenting.

	Siddiq/USA/2023 21 N = 132	Ravi/USA/ 2024 20 N = 104	Hassan/USA/2024 19 N = 21	Total N = 257 (100%)
Anterior circulationa	80	31	12	123 (47.9)
ICA	42	–	6	48 (18.7)
MCA	38	–	6	44 (17.1)
Posterior circulationb	52	66	9	127 (49.4)
Vertebral artery	44	37	5	86 (33.5)
Basilar artery	8	28	4	40 (15.6)
PCA	0	1	0	1 (0.4)
Unspecified symptomatic artery	–	7	–	7 (2.7)

ICA: internal carotid artery; MCA: middle cerebral artery; PCA: posterior cerebral artery.

^aAnterior circulation is calculated by adding the involvement of both ICA and MCA.

^bPosterior circulation is calculated by adding the involvement of PCA, vertebral and basilar arteries.

The mean (±SD) time from symptoms onset to intervention revealed a pooled estimate of 6.7 (±26.5) days (95% CI: 3.7–9.8; p < 0.001) with a significant heterogeneity across studies (I² = 79.3%, p = 0.002) (Figure 2). Periprocedural complication rates across three studies^18,19,21 with a total of 210 patients, revealed a pooled rate of 1.0% (95% CI: 0%–4.0%; p < 0.001, I² = 0%) (Supplementary Figure S2). Overall, a total of one case of hemorrhagic stroke and one death were reported, both due to intraoperative rupture of an unpredicted aneurysm.^18,21

Post-operative complications within the 30 days were demonstrated in 264 patients, with a pooled rate of 5% (95% CI: 3%–8%; p < 0.001, I² = 0%) (Supplementary Figure S3). A total of 8 (3.0%) strokes and/or deaths were reported within 30 days. Three deaths were related to ICH (two due to intraoperative rupture of an unpredicted aneurysm), one death due to in-hospital mortality with a pre-existing chronic disease, and two ipsilateral ischemic strokes were observed. TIA was recorded in 4 (1.5%) patients. Complications within 6 months were recorded for 148 patients,^18,19,21 with a pooled rate of 9% (95% CI: 5%–16%; p < 0.001, I² = 20.3%) (Supplementary Figure S4). Eight patients had TIAs, two had strokes, and two had ISR within the period of 6 months. Follow-up of 93 patients^19,20 was recorded in the period of 1 year, where 9 (9.7%) reported TIAs, 8 (8.6%) reported ISR, and 3 (3.2%) reported strokes (Table 3, Figure 3).

Table 3.

Complications (deaths, strokes, TIAs, in-stent restenosis) by study and timeframe, with sample sizes for each period.

	Siddiq/USA/2023 21	Hassan/USA/2023 18	Ravi/USA/2024 20	Hassan/USA/2024 19
Periprocedural complications; n (%) = 210 (100%)	N = 132 (62.9%)	N = 58 (27.6%)	N/R	N = 20 (9.5%)
Deaths	1 (0.75)	0	–	0
Strokes	0	1 (1.7)	–	0
TIAs	0	0	–	0
In-stent restenosis	0	0	–	N/R
30-days complications; n (%) = 264 (100%)	N = 97 (36.7%)	N = 58 (22.0%)	N = 91 (34.5%)	N = 18 (6.8%)
Deaths	2 (2.7)	2 (2.2)	2 (2.2)	0
Strokes	2 (2.7)	0	0	0
TIAs	0	0	4 (4.4)	0
In-stent restenosis	0	0	0	N/R
Complications within 6 months; n (%) = 148 (100%)	N = 91 (61.5%)	N = 41 (27.7%)	N/R	N = 16 (10.8%)
Deaths	0	0	–	0
Strokes	2 (2.1)	0	–	0
TIAs	4 (4.1)	2 (4.9)	–	2 (9.0)
In-stent restenosis	0	1 (2.4)	–	1 (4.5)
Complications within 12 months; n (%) = 93 (100%)	N/R	N/R	N = 81 (87%)	N = 12 (13.0%)
Deaths	–	–	0	0
Strokes	–	–	3 (3.7)	0
TIAs	–	–	9 (11.1)	0
In-stent restenosis	–	–	8 (9.9)	0

TIA: transient ischemic attack; N/R: not reported.

Across the included studies, assessment of ISR was conducted through both clinical and radiological follow-up, though methodologies varied. Siddiq et al. ²¹ reported clinical follow-ups in 97 patients (mean 7.4 months), angiographic follow-up in 28 patients (median 6.5 months), and radiographic follow-up (CTA/MRA) in 50 patients. During this follow-up period, six patients had symptomatic recurrent stroke/TIA; six of these had ISR on angiography and were retreated with angioplasty. Hassan et al. ¹⁸ reported 41 patients to undergo angiographic and clinical follow-ups at 6 months. Only one patient had symptomatic ISR and underwent angioplasty retreatment in the 6-month angiographic follow-up period. Ravi et al. ²⁰ assessed ISR at 1 year in 81 patients using various imaging modalities (DSA: 38.3%; CTA: 30.9%; MRA: 17.3%), identifying 8 patients with ISR. Hassan et al. ¹⁹ reported one case of ISR on 6-month conventional angiography, which was attributed to medication nonadherence.

Considering technical outcomes, Hassan et al. ¹⁸ reported that Onyx DES stenting procedures were performed unless the lesion was in target arteries with differently sized terminal ends. Hassan et al. ¹⁹ reported all 20 procedures as technically successful, with 21 stents successfully deployed across the cohort. Additionally, no access site complications were reported in Ravi et al. ²⁰

Discussion

This meta-analysis draws from four studies, including a total of 314 patients who underwent elective intracranial stenting with Onyx DES for sICAD, thus making it the largest study to monitor short- and long-term outcomes of these stents.

In our meta-analysis, the periprocedural complication rate within 3 days of the procedure was exceptionally low at 1.0%, largely due to iatrogenic intraprocedural aneurysmal rupture or early thromboembolic events. This stands in stark contrast to earlier trials. In SAMMPRIS, ⁴ the 3-day stroke or death rate in the stenting group was approximately 11%, while VISSIT ⁵ reported an even higher early complication rate of over 18%. CASSISS,²⁶ which followed a more conservative patient selection and timing protocol, showed a 2–3% rate of stroke or death within 3 days, aligning more closely with our data. The low early complication rate with Onyx DES likely reflects both improved stent flexibility and better endothelial compatibility.

At 30 days post-procedure, the cumulative complication rate in the Onyx DES group was 3.0%, still considerably safer than the 14.7% reported in the stenting groups in SAMMPRIS ⁴ and the 24.1% in VISSIT. ⁵ These earlier trials were halted or criticized for high event rates linked to poor device trackability and endothelial trauma. In contrast, CASSISS,²⁶ which was designed to address these limitations, achieved an 8% 30-day stroke or death rate. This suggests that while both the stent technology and procedural protocols matter, the Onyx DES independently offer a strong safety profile even without stringent trial exclusion criteria. By 6 months, the Onyx DES showed a 9% complication rate, including 5.4% TIAs and 3.4% ISR. The SAMMPRIS, ⁴ VISSIT ⁵ , and CASSISS²⁶ trials did not provide isolated 6-month outcomes. However, the lower mid-term complication rate in the Onyx DES group suggests that drug-eluting properties and improved stent design may delay or reduce early neointimal hyperplasia, though the benefit appears to taper over time.

At the 12-month mark, complications with Onyx DES showed a time-dependent rise, with 9.7% TIA, 8.6% ISR, and 3.2% strokes. The SAMMPRIS ⁴ trial showed a 20% rate of stroke or death and a 19.7% rate of angiographically confirmed ISR. The VISSIT ⁵ trial showed 36.2% of stroke or death, with a 26.5% rate of angiographically confirmed ISR. CASSISS²⁶ trial reported 8% rate of stroke/death and 9.5% rate of ISR. It is worth noting, however, that not all patients underwent proper follow-up imaging to check for ISR in the abovementioned trials, which makes the direct comparison limited. Overall, Onyx DES show promising durability and a relatively lower long-term complication rate, although ISR remains a concern and warrants ongoing surveillance. Additionally, the American Heart Association emphasizes that while ISR is a common complication of stenting, it has a lesser significant clinical impact as compared to periprocedural stroke and death.²⁷

Wingspan stent

The Wingspan stent is currently the only FDA-approved device used for the treatment of sICAD. It is a self-expanding stent made of nitinol, mainly used for patients with 70–99% intracranial stenosis who have failed aggressive medical therapy.^27–30 However, there have been multiple studies that have proved the high rates of procedural complications associated with this stent.^4,5,31,32

Onyx drug-eluting stent

Recently, the Onyx DES has been used off-label for sICAD. Both the Resolute Onyx and Onyx Frontier stent platforms were originally approved for coronary artery applications. These stents have a cobalt-chromium outer layer with a platinum-iridium core built on a wire leading to an improved radiopacity and deliverability. They are balloon-expandable and have thin struts, thus improving navigation in tortuous anatomy as well as reducing vessel injury. They elute Zotarolimus, which helps in reducing neointimal hyperplasia and ISR. Onyx Frontier, actually, is a refined version of Resolute Onyx with a better crossing profile, improved flexibility, thinner struts, and optimized stent delivery system. Due to these favorable properties, the Onyx DES have been adapted for off-label use in sICAD, especially in patient's refractory to medical management.^{17–21,33–35} Additionally, Onyx DES use a rapid-exchange (monorail) system compatible with standard 190 cm guidewires, simplifying device handling and reducing procedural complexity. In contrast, the Wingspan stent system requires a 300 cm exchange-length wire due to its over-the-wire design, adding technical difficulty and increasing the risk of wire-related complications. This key difference highlights the procedural advantage of Onyx DES in terms of workflow efficiency and ease of use.

Procedural and device-specific factors

Stent size selection and vessel diameter

The choice of stent diameter is critical in optimizing outcomes for sICAD. Smaller stents (<2.5 mm) may exhibit higher rates of ISR, potentially due to reduced drug delivery surface area and insufficient mechanical support within small-caliber vessels.³⁶ Conversely, oversizing the stent relative to the vessel diameter can increase the risk of vessel injury and periprocedural complications.³⁷ Therefore, careful sizing tailored to vessel dimensions is essential to minimize adverse events.

Lesion length and complexity

Longer lesions or those with significant calcification often require multiple or longer stents, which has been correlated with increased procedural difficulty and elevated rates of ISR.³⁸ Tortuous anatomy or involvement of bifurcations further complicates stent deployment, potentially compromising stent apposition and drug elution efficacy.³⁹ These factors collectively contribute to variable clinical outcomes and underscore the challenges in treating complex ICAD lesions.

ISR and reintervention risk

Complex lesions such as near-total occlusions and eccentric plaques have been associated with increased rates of ISR and the need for target lesion revascularization.⁴⁰ The antiproliferative effects of the Zotarolimus-eluting stent and prolonged dual antiplatelet therapy may mitigate this risk, but lesion morphology continues to influence long-term outcomes.

Procedural refinements

Procedural refinements play a critical role in mitigating ISR following intracranial stenting for ICAD. Pre-procedural lesion preparation, including balloon angioplasty and plaque modification, facilitates optimal stent apposition and reduces elastic recoil, which are key contributors to ISR.³⁸ Post-dilation to ensure complete stent expansion addresses malapposition and further decreases ISR risk.³⁹ Moreover, prolonged dual antiplatelet therapy contributes to preventing thrombotic events that may promote neointimal hyperplasia.³⁶ The integration of intravascular imaging modalities like intravascular ultrasound or optical coherence tomography may further optimize stent deployment and improve long-term vessel patency.⁴¹ Collectively, these procedural refinements enhance the safety and efficacy of stenting in sICAD and reduce the incidence of ISR.

Strengths

This meta-analysis stands out as the largest to date focusing on the use of Onyx DES in elective intracranial stenting for sICAD. By including four contemporary studies, the analysis reflects current trends in endovascular techniques, device usage, and operator expertise. The pooled data show a remarkably high procedural success rate of 99% and a low periprocedural complication rate of just 1%, underscoring the safety and technical reliability of Onyx DES. Another important strength is the diversity of the patient populations included, with various demographic and comorbidity backgrounds, which enhances the generalizability of the findings. The authors also took meticulous care to avoid duplication of patient cohorts—particularly between Hassan 2020 and 2023—ensuring data integrity. Furthermore, the analysis provides a temporal overview of complication rates at different follow-up intervals (3 days, 30 days, 6 months, and 12 months), giving clinicians a realistic sense of short- and intermediate-term risks associated with these devices.^{16,29,35,42,43}

Limitations

Despite its strengths, the study has a few limitations. Most notably, only four studies met the inclusion criteria, and these were primarily retrospective and single-center in design, which raises concerns about selection bias and limits the external validity of the results. Long-term follow-up beyond six months was either inconsistent or absent in some studies, making it difficult to assess the durability of the stenting outcomes over the years.

Another issue is the inconsistent patient compliance in follow-up imaging, as well as the imaging protocols. Variability in the type and frequency of imaging used to detect ISR could impact the uniformity and reliability of ISR reporting. Importantly, there were no direct head-to-head comparisons with the Wingspan stent in the included studies; thus, conclusions regarding comparative efficacy are inferred rather than explicitly demonstrated. Although funnel plots and Egger's test were used to assess publication bias, the reliability of these methods is limited due to the small number of included studies (n = 4), and thus, the risk of publication bias cannot be confidently ruled out. Additionally, critical variables such as lesion complexity, vessel anatomy, and operator experience were not uniformly controlled across studies, introducing potential confounders that could influence procedural outcomes. In addition, the device selection and stent size may have varied significantly and could influence outcomes, particularly ISR.

Clinical implications and future directions

The findings from this analysis suggest that Onyx DES may be a safer and more effective alternative to traditional stents like Wingspan for the elective treatment of sICAD, particularly in patients who have failed optimal medical therapy. The low complication rates and high procedural success make these stents highly appealing, even in anatomically complex or tortuous intracranial vessels.

Clinically, this growing body of evidence supports the integration of Onyx DES into endovascular practice, although it remains off-label for this indication. This further underscores the need for prospective, randomized controlled trials that directly compare Onyx stents to Wingspan and to aggressive medical management under standardized protocols. Future research should also extend the follow-up period to 2–5 years to provide a more comprehensive understanding of ISR, stroke recurrence, and long-term vessel patency.

Additionally, studies investigating patient-specific and lesion-specific predictors such as plaque morphology, arterial location, and degree of vessel tortuosity may help optimize patient selection and procedural planning. As the use of Onyx DES continues to rise, there is a case for pursuing regulatory approval for intracranial applications and for establishing real-world registries that monitor safety, efficacy, and rare adverse events over time.

Conclusion

In conclusion, this meta-analysis shows potential safety and durability benefits of the off-label use of Onyx DES, though further evidence is needed in treating sICAD. With significantly lower early complication rates and favorable mid-to-long-term outcomes, these stents represent a promising evolution in endovascular therapy. However, larger prospective trials are essential to validate these findings and define their long-term clinical value.

Abbreviations

Onyx DES

Onyx (Frontier and Resolute) drug-eluting balloon-mounted stents

ICAD

intracranial atherosclerotic disease

sICAD

symptomatic intracranial atherosclerotic disease

ISR

in-stent restenosis

DSA

digital subtraction angiography

CTA

computed tomography angiography

MRA

magnetic resonance angiography