Quantitative magnetic resonance angiography demonstrates persistent flow improvement following cerebrovascular stenting

Article information

Korean J Cerebrovasc Surg. 2026;.jcen.2026.E2025.12.001

Publication date (electronic) : 2026 May 27

doi : https://doi.org/10.7461/jcen.2026.E2025.12.001

Shyle H. Mehta^,1,2, Aliana Rao³, Brett Silber^1,2, Miriam Shao^1,2, Justin Turpin^1,2, Cassidy Werner^1,2, Daniel Lynch^1,2, Danielle Golub^1,2, Athos Patsalides^1,2, Henry Woo^1,2, Timothy White^1,2

¹Northwell Health, New Hyde Park, New York, USA

²Department of Neurosurgery, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Manhasset, New York, USA

³Department of Neurosurgery, Lewis Katz School of Medicine at Temple University, Philadelphia, PA, USA

Correspondence to Shyle H. Mehta Department of Neurosurgery, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell 500 Hofstra Blvd, Hempstead, NY 11549 USA Tel +1-917-355-7025 E-mail shyle.mehta@gmail.com

Received 2025 December 5; Revised 2026 April 17; Accepted 2026 May 4.

Abstract

INTRODUCTION

Endovascular stenting for cerebrovascular atherosclerotic disease remains a subject of considerable debate. Although aggressive medical therapy with dual antiplatelet agents (DAPT) is regarded as the initial standard of care, a subset of patients continues to experience recurrent ischemic events or demonstrate subclinical infarcts on imaging [21]. In this setting, stenting is an established treatment strategy and is considered the standard of care for patients who fail medical management [5]. Nonetheless, the long-term hemodynamic benefits of stenting remain insufficiently characterized.

A critical question following stenting is how best to evaluate its impact on cerebral hemodynamics. While conventional catheter angiography provides a detailed anatomic assessment at the time of the procedure, it is not always practical for serial monitoring due to its invasive nature and associated risks. Quantitative magnetic resonance angiography (qMRA) has emerged as a non-invasive and reproducible technique capable of objectively measuring volumetric blood flow [3,4]. This modality allows for evaluation of both immediate and durable changes in cerebral circulation after stenting without the need for repeated catheter-based studies.

Device selection also remains an area of uncertainty in the management of intracranial atherosclerotic disease. A variety of stent platforms have been utilized, yet no single device has emerged as the consensus standard. Our institutional experience suggests that the Onyx drug-eluting stent (DES) offers a safe and effective option in this context [17,18,22]. Building on our prior work, we sought to examine whether Onyx stenting leads to measurable improvements in flow that persist on long-term follow-up by comparing pre- and post-stenting qMRA flow values.

MATERIALS AND METHODS

A retrospective chart review was performed for all patients who underwent cerebrovascular stenting for atherosclerotic lesions using Onyx Frontier^TM and Resolute Onyx^TM DES (Medtronic, Minneapolis, MN, USA) at a single institution between January 2018 and September 2024. The study protocol was approved by the institutional review board, and informed consent was waived due to the retrospective design. Data collected included demographics, symptoms at presentation and follow-up, radiographic findings at presentation and follow-up (computed tomography [CT] angiography, magnetic resonance [MR] angiography, and/or digital subtraction angiography [DSA]), procedural details, and complications. Outcomes assessed were procedural success, periprocedural stroke, in-stent restenosis (ISR) requiring intervention, and qMRA flow measurements obtained pre-stenting, post-stenting (one to three days per institutional protocol), and at long-term follow-up (median of 8.9 months following intervention).

Patients were included if they had symptomatic cerebrovascular atherosclerotic disease refractory to medical therapy (i.e., dual antiplatelet therapy) and underwent stenting with Onyx DES. Stenosis severity was determined based on angiographic findings in conjunction with clinical presentation, including recurrent ischemic symptoms despite dual antiplatelet therapy. Lesions were considered moderate to severe based on the treating neurointerventionalist’s assessment, taking into account both luminal narrowing and clinical context. Exclusion criteria included non-atherosclerotic lesions, insufficient imaging follow-up, or lack of qMRA data. A representative case of stenosis successfully treated with Onyx DES is shown in Fig. 1.

Fig. 1.

Representative case demonstrating angiographic and quantitative magnetic resonance angiography (qMRA) flow changes following right vertebral artery stenting. qMRA measurements obtained pre-stenting (A), on post-operative day 1 (B), and at long-term follow-up 13 months after intervention (C) demonstrate marked immediate improvement with attenuated but persistent flow augmentation over time (pre-stent: 15 mL/min; post-operative day 1: 168 mL/min; long-term follow-up: 112 mL/min). Corresponding diagnostic cerebral angiography demonstrates right V4 segment stenosis prior to intervention (D; blue arrows) and marked luminal improvement immediately following stent placement (E). No long-term angiographic follow-up was obtained. The left vertebral artery terminates at the level of C2 without intracranial contribution; accordingly, restoration of right vertebral artery flow results in improved basilar artery flow.

Statistical analyses for descriptive statistics included summations, means, medians, and modes. Given the non-normal distribution and high variability of qMRA percent increase values, nonparametric statistical tests were used. Within-vessel changes relative to baseline were assessed with the Wilcoxon signed-rank test. Paired post-stent and long-term differences were assessed with the Wilcoxon signed-rank test. For completeness, parametric t-tests were also performed, and results were consistent with the nonparametric analyses. Analyses were conducted using R (version 4.5.0).

qMRA was performed using Non-invasive Optimal Vessel Analysis (NOVA) software with non-contrast MR imaging sequences for volumetric flow quantification. Imaging was obtained on a 1.5T scanner using standardized acquisition parameters. Flow measurements were calculated using dedicated NOVA software, which enables vessel-specific volumetric flow quantification (mL/min) through user-defined regions of interest placed orthogonal to the vessel axis.

RESULTS

Demographics

A total of 63 patients with 84 stents were included, with a mean age of 63.1±11.4 years; 33.3% were female. The median number of stents deployed per patient was 1.0±0.7, with a median stent diameter of 2.5±0.9 mm and a mean stent length of 11.0±3.9 mm. The mean proximal vessel size was 3.4±1.8 mm, and the mean percent stenosis was 77.4±11.8% (Table 1). Recurrent stenosis occurred in 15.9%, post-stent angioplasty in 6.4%, technical complications in 4.8%, and clinical complications (i.e., stent-related strokes) in 4.8% of patients. Technical complications were defined as any intra-procedural event that deviated from the expected course, including device-related issues, difficulty with deployment, or vessel injury. One patient experienced a delayed recurrent stroke despite uneventful stent placement and the absence of ISR on follow-up angiography. This patient demonstrated persistently low flow on post-stent qMRA, with values below the age-adjusted standardized normal range for the treated vascular territory. Long-term angioplasty was required in 4.8% of cases and was defined as angioplasty performed at least 3 months after the index procedure. Regarding stented vessels, the most common locations were the basilar artery (BA) (22.6%), middle cerebral artery (MCA) (20.2%), and V4 vertebral artery (VA) (16%), followed by the supraclinoid internal carotid artery (ICA) (10.7%), cavernous ICA (8.3%), petrous ICA (4.8%), cervical ICA (2.4%), V1 (10.7%), and V3 (1.2%). The median time to long-term qMRA follow-up was 9.4 months (range, 3.3-52.2 months).

Table 1.

Demographics. Long-term stent angioplasty refers to angioplasty performed within 2.9–21.1 months (median 9.9 months (IQR 4.0–17.4)) following index stent placement

qMRA changes over time

qMRA flow volumes of the stented vessel were recorded for the entire cohort pre-stent, post-stent, and, if possible, at long-term follow-up. Immediately following stent placement, patients demonstrated substantial increases in flow with a broad distribution and several high-value outliers. The median percent flow increase was 150.8% (IQR 71.4-347.9). At long-term follow-up, the distribution shifted toward lower values, although flow remained elevated in most patients, with a median percent increase of 73.3% (IQR 24.0-195.2) (Fig. 2). Comparison of paired measurements demonstrated a significant reduction in percent flow increase over time (Wilcoxon signed-rank test, p=0.043). These findings indicate that stenting produces a significant and durable enhancement to flow relative to pre-stent values, with only modest attenuation over time.

Fig. 2.

Quantitative magnetic resonance angiography (qMRA) percent flow increase following stent placement and at long-term follow-up. Violin plots illustrate the distribution of percent flow increase measured by qMRA immediately post-stent and at long-term follow-up. Individual patient values are shown as black dots. The embedded boxplots represent the interquartile range with the median indicated by the horizontal line, while the whiskers denote the overall data spread. Red dots indicate the mean percent flow increase for each time point. The y-axis is displayed on a logarithmic scale to accommodate the wide range of observed flow changes.

Absolute cerebral blood flow values at follow-up demonstrated substantial variation across vascular territories. The ICA exhibited the highest mean flow (148.59±104.49 mL/min), followed by the basilar artery (84.92±45.33 mL/min) and vertebral artery (83.95 ± 57.39 mL/min), while the MCA demonstrated lower absolute flow values (48.91±32.12 mL/min)(Table 2). These findings correspond to relative flow increases of 142.71% in the ICA, 84.92% in the basilar artery, 63.41% in the vertebral artery, and 39.8% in the MCA, consistent with vessel-specific hemodynamic augmentation following intervention.

Table 2.

Absolute flow values by vessel

When stratified by vessel, all arteries demonstrated robust increases in qMRA percent change immediately after stenting (Fig. 3). The ICA exhibited the largest median increase post-stent at 178.3% (IQR 77.0-347.9), followed by the VA at 157.1% (IQR 47.6-1020.0), the basilar artery at 154.5% (IQR 115.6-359.8), and the MCA at 45.7% (IQR −4.6-94.6). At long-term follow-up, median flow augmentation remained elevated in the VA (130.1%, IQR 73.3-646.7), while the basilar (75.9%, IQR 49.7-520.3), MCA (50.6%, IQR 36.7-123.2), and ICA (38.0%, IQR 10.9-72.5) demonstrated attenuation of augmentation. Significant reductions in flow augmentation over time were observed in the VA (p=0.036) and ICA (p=0.025), while changes in the basilar (p=1.00) and MCA (p=0.063) were not statistically significant.

Fig. 3.

Distribution of quantitative magnetic resonance angiography (qMRA) percent flow increase following intracranial stent placement by vessel type. Boxplots demonstrate the percent increase in cerebral blood flow measured by qMRA immediately post-stent placement and at long-term follow-up across four vascular territories: basilar artery (Basilar), vertebral artery (VA), middle cerebral artery (MCA), and internal carotid artery (ICA). The central horizontal line represents the median, the box denotes the interquartile range, and the whiskers indicate the range of observed values. Across vessel types, post-stent measurements generally demonstrate greater flow augmentation compared with long-term follow-up, with variability observed between vascular territories.

The basilar artery showed a mean increase of ~338% post-stent (median 133%, Wilcoxon p=0.0019), which persisted with some attenuation at long-term follow-up (~294%, median 76%, p=0.029). The VA exhibited the greatest augmentation in flow, with mean post-stent increases exceeding 631% (median 125%, p<0.001), which was attenuated but persistent at long-term (~453%, median 145%, p=0.0016). The MCA showed more modest increases (mean ~107% post-stent, median 106%, p=0.0122), but values remained significantly above baseline at long-term (~84%, median 50.6%, p=0.0295). The ICA displayed large but highly variable increases post-stent (mean ~651%, median 196%, p<0.001), which persisted with some attenuation at long-term follow-up (mean ~417%, median 38%, p=0.0105). These results confirm that all major cerebrovascular vessels experience durable and significant flow augmentation following stenting (Table 3).

Table 3.

Within-vessel vs. 0%. Long-term stent angioplasty refers to angioplasty performed within 2.9–21.1 months (median 9.9 months (IQR 4.0–17.4)) following index stent placement

When comparing matched post-stent and long-term measurements, most vessels demonstrated stable augmentation. The ICA demonstrated a statistically significant decline (ΔMedian -140.3%, p=0.0253), indicating that ICA flow augmentation diminishes over time, while other vessels maintain more stable changes. The BA showed a decline from post-stent to long-term (ΔMedian -100%), but this change was not significant. The VA showed a small decline (ΔMedian -27%), which reached significance (p=0.0173). Importantly, the MCA showed a slight increase (ΔMedian +4.9%), which approached significance (p=0.0591) (Table 4).

Table 4.

Paired post-stent vs. Long-term (within vessel)

When all vessel territories were analyzed together, 36 paired measurements were available. The median percent flow increase decreased from 152% immediately post-stent to 72.9% at long-term follow-up. Paired Wilcoxon signed-rank testing demonstrated a significant reduction in percent flow increase between the post-stent and long-term measurements (p=0.016).

In-stent restenosis

A total of 11 vessels in 10 patients demonstrated angiographic evidence of ISR. These included 8 anterior circulation segments—supraclinoid ICA (n=4), cavernous ICA (n=2), and MCA (n=2)—and 3 posterior circulation segments (V4=2, BA=1). Of these patients, three (4.8%) underwent delayed angioplasty. Two cases were identified on routine follow-up angiography, while one was prompted by new neurological symptoms. The remaining cases were managed conservatively, as they were radiographic rather than symptomatic. All vessels with ISR demonstrated a decline in qMRA flow at long-term follow-up. The mean change in flow relative to post-stent measurements was −127.5±75.7 mL/min, with a median change of −107.5 mL/min. The mean percent flow decline was 54.7%, with a median decline of 52.7%. A representative case of ISR after treatment with Onyx DES is shown in Fig. 4.

Fig. 4.

Representative case demonstrating in-stent restenosis (ISR) following tandem stenting of the left cavernous internal carotid artery and left M1 segment, with corresponding quantitative magnetic resonance angiography (qMRA) flow changes. Pre-stent diagnostic cerebral angiography (A) demonstrates high-grade stenosis of the left M1 segment (blue arrow), with corresponding reduced flow on qMRA (B; 39 mL/min). Post-stent angiography (C) shows marked luminal improvement with improved filling of the left middle cerebral artery (MCA) territory, accompanied by increased flow on qMRA obtained on post-operative day 1 (D; 116 mL/min). Follow-up angiography at 1 year (E) demonstrates mild ISR of the left M1 segment (blue arrow), with preserved improvement in caliber and MCA territory perfusion compared to pre-stenting. Long-term qMRA at 5 months (F) demonstrates susceptibility-related signal dropout at the stented left M1 segment, with decreased flow in the distal M2 branches relative to immediate post-procedural imaging, though still improved compared to pre-stenting.

DISCUSSION

Endovascular stenting for cerebrovascular atherosclerotic disease remains an area of active debate. Although DAPT is the recommended first-line management, a substantial subset of patients continues to experience recurrent ischemic events despite adherence [9,16]. For these patients, stenting represents an established therapeutic option [1]. However, evaluating its long-term hemodynamic benefit has been limited by the lack of reliable, non-invasive follow-up tools.

qMRA provides a reproducible, non-invasive method of measuring vessel-specific volumetric cerebral blood flow and addresses this gap. Prior studies have shown its ability to detect hemodynamic compromise before clinical deterioration, and its utility has been demonstrated in moyamoya disease and vertebrobasilar [2,3,10]. Our findings extend this role by demonstrating robust increases in cerebral blood flow following Onyx stent deployment, with augmentation persisting, with some attenuation, at long-term follow-up. These results highlight qMRA’s value not only as a tool for documenting procedural success but also for non-invasively monitoring ISR and guiding the need for confirmatory angiography, with a primary focus on hemodynamic endpoints rather than clinical outcomes.

Importantly, flow augmentation was observed across all major vascular territories. Both anterior and posterior circulation vessels demonstrated durable increases, although the ICA and VA showed significant attenuation in flow over time. This attenuation likely reflects progressive hemodynamic compromise from developing ISR, consistent with prior reports identifying these territories as particularly prone to ISR [8,12]. In our cohort, all patients with angiographically confirmed ISR exhibited measurable qMRA flow declines at long-term follow-up (median 9.4 months), with a mean reduction of approximately 55% from post-stent levels. These findings demonstrate a strong correspondence between hemodynamic attenuation and angiographically confirmed ISR, supporting qMRA as a sensitive, non-invasive surveillance tool capable of detecting evolving ISR before symptomatic presentation. The mechanisms underlying these changes are likely multifactorial—including chronic inflammation, neointimal hyperplasia, stent under-expansion, and patient comorbidities—and warrant further investigation.

The magnitude of percentage-based flow augmentation observed in this study should be interpreted with caution. Large relative increases, particularly in the ICA and vertebral artery territories, likely reflect low baseline flow states rather than supraphysiologic post-intervention flow. Accordingly, both absolute and relative flow metrics are presented to provide a more physiologically meaningful interpretation. Importantly, this study focuses on hemodynamic endpoints as measured by qMRA. Clinical outcomes following Onyx stenting in this cohort have been previously reported, demonstrating high procedural success, low complication rates, and low rates of reintervention [21]. The present analysis was designed to provide complementary insight into the hemodynamic effects of stenting rather than to re-evaluate clinical outcomes. While flow augmentation is hypothesized to correlate with improved clinical outcomes, future studies integrating both hemodynamic and clinical endpoints will be essential to define this relationship.

The safety profile of Onyx stenting in our series was favorable. Technical complication and clinical complication rates were each 4.8%, consistent with or below prior reports (<10%) [14,23]. One patient experienced delayed recurrent stroke despite uneventful stent placement and absence of ISR on follow-up angiography. Post-stent qMRA in this case demonstrated persistently low flow, with values below the age-adjusted standardized normal range for the treated vascular territory. This finding suggests that recurrent stroke may have resulted from insufficient hemodynamic augmentation or from progressive atherosclerotic disease with ongoing distal embolization. ISR was observed in a subset of patients and was consistently associated with qMRA-derived flow attenuation at follow-up. While radiographic restenosis was not uncommon, clinically significant cases requiring reintervention were relatively infrequent, supporting the utility of qMRA as a non-invasive tool for hemodynamic surveillance.

Our study also supports emerging evidence for the role of DES in cerebrovascular disease. Onyx DES offer several advantages in the neurovascular setting, including low-profile delivery systems, enhanced navigability in tortuous anatomy, and greater radial force compared with self-expanding stents. Additionally, the antiproliferative drug coating may reduce neointimal hyperplasia and subsequent ISR. However, their use in cerebrovascular lesions remains off-label, and careful patient selection is essential. In the coronary literature, Onyx DES have been shown to reduce ISR in small vessels [20]. Recent cerebrovascular series similarly demonstrate the safety and efficacy of Onyx DES, as well as superiority compared with bare-metal stents [6,11,13]. Our findings contribute to this growing body of data, demonstrating persistent hemodynamic benefit and low complication rates. Compared with alternatives such as the Wingspan Stent System, which was associated with higher ISR and adverse outcomes in SAMMPRIS, Onyx DES appear to offer improved navigability, deployment characteristics, and long-term durability [7,15].

This study has several limitations. Its retrospective design precludes causal inference, and follow-up qMRA was not available in all patients, introducing potential selection bias. Although qMRA is reproducible, measurements may still be influenced by systemic hemodynamics, operator technique, or cardiac output variability [3,4,19]. Variability in the timing of post-stent qMRA acquisition may influence measured flow changes, as early imaging may capture transient hyperemia rather than stable hemodynamic remodeling. Additionally, subgroup analyses—particularly for basilar and MCA territories at long-term follow-up—were limited by small sample sizes and should be interpreted with caution. Comparisons across vascular territories did not account for within-subject correlation in patients contributing multiple vessels, which may affect the precision of between-group comparisons. Furthermore, heterogeneity in lesion morphology and stent characteristics may have contributed to variability in flow responses. The inclusion of multiple vascular territories and segments introduces additional heterogeneity, which may limit the generalizability of these findings to specific vascular distributions. Future prospective multicenter studies with standardized imaging intervals and integrated clinical outcome analysis will be essential to better define the durability of flow augmentation and identify patient subgroups most likely to benefit.

In summary, our study demonstrates that Onyx drug-eluting stents provide durable hemodynamic improvements across both intracranial and extracranial cerebrovascular vessels, as measured by qMRA. These findings support qMRA as a valuable adjunct to traditional catheter angiography for post-stent follow-up and suggest that Onyx stents represent a safe and effective option for patients with cerebrovascular atherosclerotic disease refractory to medical therapy. Larger prospective studies are warranted to validate these results and refine patient selection.

CONCLUSIONS

Onyx DES provide robust and durable augmentation of cerebral blood flow across both intracranial and extracranial segments supplying the brain, as demonstrated by qMRA. Flow improvements were persistent with some attenuation over time across both anterior and posterior circulations, though the ICA and vertebral arteries exhibited some attenuation over time. These findings establish qMRA as a practical, non-invasive method for monitoring hemodynamic outcomes and detecting ISR. The favorable safety profile and durable flow augmentation support the role of Onyx stents as an effective option for patients who remain symptomatic despite medical therapy. Prospective, multicenter studies are warranted to confirm these findings and refine patient selection.

Article information Continued