Preoperative embolization of feeding arteries in glioblastoma: Technical strategies and clinical utility

Abstract

Background

Maximal safe resection—balancing aggressive tumor removal with neurological preservation—is essential in glioblastoma surgery. Cerebral angiography enables identification of the dominant hemisphere, feeding arteries, and vessels crossing eloquent areas, thereby aiding intraoperative planning. Preoperative embolization of glioblastoma feeders may shorten operative time and reduce intraoperative bleeding, similar to its established role in extra-axial tumors. We hypothesized that embolization could serve not as a routine adjunct but as a targeted strategy for selected glioblastomas with deep-seated or hypervascular feeders that increase surgical complexity.

Methods

Between December 2023 and July 2025, 15 consecutive patients with glioblastoma underwent preoperative embolization under local anesthesia using a 3-Fr distal radial approach. Coils and n-butyl-2-cyanoacrylate (NBCA) were used as embolic materials. Cerebral angiography and the Wada test (intracarotid amobarbital procedure) were performed to evaluate vascular anatomy and hemispheric dominance.

Results

The embolization procedure required a mean of 1 h 26 min. Microscope-assisted tumor resection averaged 3 h 1 min, with a mean blood loss of 389 mL. No neurological or ischemic complications occurred, and postoperative MRI confirmed the absence of new ischemic changes.

Conclusions

Preoperative embolization of glioblastoma-feeding arteries may provide a clear and bloodless surgical field, enhances spatial orientation through radiopaque contrast marking, and facilitates precise intraoperative localization. By reducing intraoperative bleeding, surgical complexity, and anesthesia time, this selective technique may decrease the overall invasiveness of glioblastoma surgery. When performed via a minimally invasive distal radial approach, preoperative embolization is a safe, feasible, and practical adjunct that enhances the precision and efficiency of glioblastoma resection.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00701-026-06796-5.

Introduction

Glioblastoma is the most aggressive primary brain tumor, and despite multimodal therapy, median survival remains 15–18 months [20]. Achieving maximal safe resection is a major determinant of outcome [4, 17, 21], and detailed preoperative evaluation of vascular anatomy can facilitate this goal—particularly in tumors with deep or hypervascular components. Cerebral angiography allows identification of dominant feeders, transit arteries, and vessels crossing eloquent regions, information that cannot be fully obtained from CT or MR angiography.

At our institution, angiography is selectively incorporated into the preoperative workup for glioblastomas with suspected hypervascularity or complex vascular relationships. When feeders arise close to eloquent cortex or contribute substantially to tumor perfusion, the intracarotid amobarbital (Wada) test is performed concurrently to assess hemispheric dominance. In selected cases, preoperative embolization of tumor-feeding arteries is undertaken during the same session to reduce intraoperative bleeding, shorten operative time, and improve anatomic orientation through the use of radiopaque embolic materials as intraoperative markers.

Preoperative embolization is widely used in meningioma surgery, but its application to glioblastoma has been limited to isolated case reports and small series, and indications remain poorly defined. The present study describes our consecutive institutional experience with this strategy, focusing on its technical feasibility, safety considerations, and intraoperative utility in appropriately selected glioblastoma cases.

Methods

Study design and ethical approval

This single-institution retrospective study was conducted at our tertiary referral center between December 2023 and July 2025. The protocol was approved by the institutional ethics committee (IRB No. 7216) and complied with the Declaration of Helsinki. Written informed consent for diagnostic angiography, preoperative embolization, and tumor resection was obtained from all patients.

Safety and technical considerations

Embolization of pial or perforator feeders is associated with a well-recognized risk of ischemic injury to eloquent cortex, including inadvertent territorial infarction, reflux-related injury, and delayed ischemic complications [8, 22]. In recognition of these risks, our embolization strategy was deliberately conservative and safety-oriented.

Embolization was performed only when tumor-restricted perfusion could be unequivocally confirmed by detailed super-selective angiography and continuous neurological monitoring under local anesthesia. Feeders suspected to supply eloquent cortex, to function as en passage arteries, or to give rise to critical perforators were strictly excluded from embolization. Distal parenchymal embolization was intentionally avoided, and embolization was limited to proximal flow reduction to minimize the risk of reflux or unintended ischemia.

A standardized protocol was followed, including systemic heparinization (activated clotting time > 250 s), continuous catheter flushing with heparinized saline, and super-selective test injections to confirm tumor-restricted perfusion of the targeted feeder.

A 3-Fr distal radial approach was used in all cases, and low-profile microcatheters were advanced into distal cortical or perforating branches. Depending on vessel caliber and flow characteristics, selective feeder occlusion was achieved using detachable coils or small-volume NBCA (1:3 Lipiodol dilution), which also served as radiopaque intraoperative markers during tumor resection (Figs. 1, 2 and 3).

Fig. 1.

Preoperative embolization and intraoperative findings in a left temporal glioblastoma (Case 6). A Contrast-enhanced T1-weighted MRI showing a hypervascular tumor in the left temporal lobe. B Left ICA angiography demonstrating intense tumor staining supplied by the early temporal artery (ETA). C Intraoperative angiographic image showing coil deployment within the ETA. D Post-embolization angiography confirming complete devascularization and disappearance of tumor staining. E Distal transradial access using a 3-Fr guiding sheath. F After Sylvian fissure dissection, the embolized coil served as an immediate intraoperative landmark for recognizing the ETA. G The superficial middle cerebral vein (SMCV) crossing the tumor surface was preserved; venous peeling was bloodless due to prior feeder control. H Tumor resection proceeded along the gliosis–tumor interface with excellent hemostasis. I Intraoperative view after complete tumor removal. J Postoperative contrast-enhanced T1-weighted MRI confirming gross-total resection

Fig. 2.

Preoperative embolization and intraoperative findings in a right frontal glioblastoma (Case 12). A Contrast-enhanced T1-weighted MRI showing a right frontal tumor extending toward the insula. B Lateral ICA angiogram revealing tumor supply from branches of the callosomarginal artery. C Selective angiography showing feeders from the anterior internal frontal artery (AIFA) and the anterior branch of the middle internal frontal artery (MIFA), with the posterior MIFA branch identified as a passing artery. D Intraoperative angiographic image showing coil embolization of the AIFA and anterior MIFA branch. E Post-embolization angiography demonstrating complete elimination of tumor staining. F During interhemispheric dissection, the coil marker in the anterior MIFA branch facilitated identification of the posterior MIFA branch and the posterior internal frontal artery (PIFA). G Temporary clipping confirmed the functional significance of the posterior MIFA branch, which was preserved. H Tumor removal along the gliosis–tumor border proceeded under excellent hemostasis. I Intraoperative field after complete resection. J Postoperative MRI confirming total resection

Fig. 3.

Visualization of embolized feeders by NBCA as intraoperative landmarks in a left frontal glioblastoma (Case 1). A After interhemispheric fissure dissection, an NBCA cast (asterisk) marking the anterior internal frontal artery (AIFA) was clearly visualized, enabling rapid intraoperative identification of the feeder. B The NBCA cast also facilitated delineation of adjacent arterial anatomy, improving depth perception and orientation during resection. C Preoperative contrast-enhanced T1-weighted MRI demonstrating a left frontal glioblastoma with deep arterial supply. D Postoperative contrast-enhanced T1-weighted MRI confirming complete tumor removal. This case highlights the utility of NBCA as a radiopaque anatomical marker that enhances intraoperative orientation in glioblastoma surgery

After each procedure, access-site hemostasis was obtained manually, patients were monitored neurologically for at least 24 h, and postoperative MRI was performed to exclude new ischemic lesions. This protocol is consistent with established safety benchmarks for diagnostic and interventional cerebral angiography, which report neurological complication rates of 2–3% and permanent deficits < 0.2% [7, 12].

Patient selection

Patients with newly diagnosed or recurrent glioblastoma were screened for angiographic evaluation before surgery. Candidates for preoperative embolization included patients with deeply located or hypervascular tumors demonstrating identifiable pial or perforating feeders amenable to super-selective catheterization. Importantly, case selection was based on qualitative angiographic and anatomical assessment rather than predefined quantitative thresholds. Specifically, embolization was considered only when a surgically relevant feeder could be clearly identified, tumor-restricted perfusion was confirmed on super-selective angiography, and the physiological territory of the vessel was judged to be limited.

Feeders suspected to supply eloquent cortex, to function as en passage arteries, or to give rise to critical perforators were deliberately excluded from embolization.

In recurrent tumors with post-treatment fibrosis or deep feeders, embolization was performed to re-establish a clean dissection plane and reduce intraoperative congestion. Exclusion criteria included diffuse or poorly defined vascular supply, feeders arising from critical perforators supplying eloquent cortex, and contraindications to angiography such as renal dysfunction or contrast allergy.

Pre-angiographic evaluation

Preoperative gadolinium-enhanced MRI, diffusion tensor imaging, and CT angiography were used to delineate tumor anatomy and plan the angiographic approach. Cerebral DSA was performed to map arterial and venous anatomy, identify dominant feeders, and assess vascular displacement.

The intracarotid amobarbital procedure (Wada test) was performed when clarification of hemispheric dominance or language localization was necessary, particularly when intended embolization was near eloquent regions. This practice aligns with established pre-surgical functional lateralization standards [1].

The procedure was conducted in 4 of 15 patients (27%) when functional dominance could not be definitively established on noninvasive imaging.

Endovascular procedure

A 3-Fr Axcelguide Stiff-J guiding sheath (Medikit, Tokyo, Japan) was introduced via the distal radial artery and advanced to the target internal carotid or vertebral artery. Super-selective catheterization was performed using an Excelsior SL-10 microcatheter (Stryker, Kalamazoo, MI, USA) over a Synchro Select 0.014-inch guidewire.

Tumor-feeding arteries were identified by characteristic delayed, dense parenchymal staining without supply to normal cortex.

Selective embolization was performed using detachable coils or NBCA, depending on vessel caliber and flow dynamics, as illustrated in Fig. 3. The choice of embolic material was individualized according to vessel caliber, flow dynamics, and anatomical configuration, reflecting real-world procedural variability rather than a protocolized material-specific strategy. Complete devascularization was confirmed angiographically. Embolization was intentionally limited to vessels in which functional safety could be confidently ensured. Feeders suspected to supply normal cortex or to function as en passage arteries were deliberately excluded. This strategy aimed to balance effective flow reduction with preservation of functional cortex rather than achieving angiographic completeness.

Embolization was typically performed the day before tumor resection to maximize hemostatic benefit while minimizing recanalization and histopathological alteration [2].

Timing of surgery

Tumor resection was scheduled within 24 h after embolization to preserve flow-arrest effects and minimize the likelihood of feeder recanalization. When logistical constraints necessitated delay, surgery was performed within 36 h.

Surgical procedure

All tumor resections were performed under a surgical microscope using neuronavigation and intraoperative neurophysiological monitoring.

Preoperative feeder occlusion resulted in reduced tumor surface congestion, decreased bleeding during cortical entry, and improved anatomical orientation throughout the operation.

Embolized coils (Figs. 1 and 2) and NBCA casts (Fig. 3) provided radiopaque and tactile intraoperative landmarks that facilitated identification of feeder origins and resection depth, consistent with prior reports demonstrating improved orientation and reduced bleeding after feeder embolization [10, 16, 18].

These benefits allowed precise dissection along the tumor–brain interface and minimized the need for bipolar coagulation.

Microscope-assisted operative time and estimated blood loss were prospectively recorded.

Data collection and analysis

Demographic information, tumor characteristics, angiographic findings, embolic materials, and procedural parameters were collected. Intraoperative variables—including microscope-assisted operative time (from dural opening to completion), estimated blood loss, and anesthesia duration—were documented by the anesthesia team.

Postoperative MRI within 72 h was used to assess extent of resection and identify ischemic changes. Continuous variables are presented as mean ± standard deviation and categorical variables as counts and percentages. Given the exploratory nature of this single-arm study, analyses were descriptive without formal statistical testing.

Results

From December 2023 to July 2025, 15 patients with glioblastoma underwent preoperative feeding-artery embolization followed by microsurgical resection.

Patient characteristics—including sex, age, tumor location and side, punctured vessel, microcatheter type, embolization material, and procedure time—are summarized in Table 1.

Table 1.

Summary of 15 consecutive cases undergoing preoperative embolization for glioblastoma

Case	Age	Sex	Tumor location	Approach	Feeding artery	Microcatheter	Embolic material	Operative time (h:mm)	Blood loss (mL)
1	70	F	Lt frontal	Radial	AIFA, MIFA	SL-10	NBCA 20%,15%,coil	2:05	867
2	75	M	Lt occipital	Distal radial	IOA (parieto-occipital br., calcarine br.)	SL-10	NBCA 30%, coil	3:54	185
3	49	M	Rt temporal	Radial	Anterior temporal a	SL-10	Coil	2:09	357
4	70	M	Rt frontal	Distal radial	Precentral a	SL-10	Coil	2:35	80
5	46	M	Rt occipital	Distal radial	Calcarine a	SL-10	Coil	2:22	295
6	71	M	Lt temporal	Distal radial	Early temporal a	SL-10	Coil	2:54	84
7	49	M	Lt temporal	Distal radial	Anterior temporal a	SL-10	Coil	4:27	618
8	68	F	Rt temporal	Distal radial	Anterior temporal a	MARVEL 1.9 Fr	Coil	2:18	159
9	75	F	Lt frontal	Distal radial	AIFA, MIFA	MARVEL 1.9 Fr	Coil	2:42	701
10	63	M	Lt temporal	Distal radial	Anterior & middle temporal a	SL-10	Coil	2:51	990
11	74	F	Rt temporal	Distal radial	Posterior temporal a	Marathon	NBCA 30%	1:04	130
12	63	M	Rt frontal	Distal radial	AIFA, MIFA	SL-10	Coil	3:50	128
13	83	M	Rt temporal	Distal radial	Middle temporal a	SL-10	Coil	2:56	136
14	36	M	Rt thalamus	Distal radial	Lateral & medial PChoA	DeFrictor NANO	NBCA 30%	4:09	685
15	61	M	Lt temporal	Distal radial	Early temporal a	SL-10	Coil	5:11	427

Patient characteristics including age, sex, tumor location and side, punctured vessel, microcatheter used, embolic material, and procedural time

ETA early temporal artery, MIFA middle inferior frontal artery, AIFA anterior inferior frontal artery, SMCV superior middle cerebral vein, IAP intracarotid amobarbital procedure, IOA internal occipital artery, PchoA posterior choroidal artery, D distal

Units: h = hours; min = minutes; mL = milliliters

• Approach: Radial = conventional radial approach; Distal radial approach = dTRA

• AIFA anterior internal frontal artery, MIFA middle internal frontal artery, IOA internal occipital artery, PChoA posterior choroidal artery

• Operative time = microscope-assisted time from dural opening to completion

• Two cases (Case 7 and Case 15) were performed under awake surgery

The mean duration of the embolization procedure, including vascular mapping and the intracarotid amobarbital procedure (Wada test, IAP), was 1 h 26 min.

The mean microscope-assisted operative time was 3 h 01 min, and the mean blood loss was 389 mL.

Excluding the two awake surgeries, the remaining 13 cases had an average operative time of 2 h 45 min and a mean blood loss of 369 mL.

No procedure-related complications occurred, including puncture-site events. Postoperative MRI demonstrated no ischemic changes beyond the resection margins in any case. Although MRI was not obtained between embolization and tumor resection, no macroscopic findings suggestive of ischemic tissue—such as discoloration, tissue softening, or indistinct cortical–subcortical boundaries—were observed intraoperatively.

The intraoperative findings described in the Surgical procedure section were consistently reproduced in all patients—specifically, reduced intraoperative bleeding, clearer anatomical orientation, and the utility of embolic materials as surgical landmarks.

Gross-total or near-total resection was achieved in all patients without new neurological deficits.

Illustrative angiographic and operative details are presented in Case 6 and Case 12 (Figs. 1 and 2), which exemplify the technical strategies and intraoperative benefits described above.

Llustrative cases

Representative examples of our technique are presented in Figs. 1, 2, and 3 and Supplementary Videos 1–2. These cases highlight the key intraoperative advantages of targeted feeder embolization, including reduced arterial inflow, minimized venous congestion, and improved surgical orientation through radiopaque coil or NBCA markers. Detailed angiographic and operative sequences are available in the accompanying figure panels and video materials.

Discussion

The present study should be interpreted in light of its retrospective, single-arm design and limited sample size. Importantly, this investigation was not intended to establish a causal relationship between preoperative embolization and reductions in intraoperative blood loss or operative time. Rather, our primary objective was to evaluate the technical feasibility, procedural safety, and intraoperative utility of selective feeder embolization in carefully selected glioblastoma cases. Because no contemporaneous non-embolized control group was included, the observed intraoperative findings may have been influenced by multiple confounding factors, including tumor biology, anatomical complexity, surgeon experience, and perioperative management. Accordingly, any associations observed in this series should be regarded as descriptive rather than evidence of comparative effectiveness.

The present study evaluated the technical feasibility, safety profile, and intraoperative utility of preoperative feeder embolization for glioblastoma. Our study provides several novel findings. First, to our knowledge, this is the first consecutive series in which preoperative embolization for glioblastoma was performed exclusively through a 3-Fr distal transradial approach, demonstrating that this minimally invasive access is feasible even for super-selective catheterization of pial arterial feeders. Second, we show that coils and NBCA casts can function as intentional intraoperative radiopaque landmarks, providing consistent spatial orientation in deep or distorted anatomy—an application that has not been systematically described in prior glioblastoma literature. Third, we propose practical angiographic selection criteria based on feeder anatomy and physiological territory, addressing a major gap in current evidence where indications for glioblastoma embolization remain undefined. Finally, the reproducible reduction in intraoperative bleeding and the absence of ischemic complications across all patients provide supportive evidence for the procedural safety and potential utility of this strategy in appropriately selected cases.

Significance and utility of angiography

Maximal safe resection of glioblastoma requires precise understanding of arterial inflow and venous drainage patterns. Although CT and MR angiography provide useful anatomical detail, they cannot reliably distinguish tumor-specific perfusion from normal cortical supply. Cerebral digital subtraction angiography (DSA) remains indispensable for identifying selective tumor-feeding arteries, transit arteries, and venous displacement associated with mass effect or edema (illustrated in Figs. 1B and 2B–C). Perfusion MRI complements DSA but only angiography provides dynamic information directly applicable to endovascular planning [6, 14].

In this context, selective preoperative angiography enables detailed vascular interpretation and informs the feasibility and safety of embolization in carefully selected glioblastoma cases.

Endovascular approaches in glioblastoma

The existing literature on preoperative embolization for glioblastoma is limited and consists primarily of isolated case reports and small series, often involving highly hypervascular tumors or AVM-like angioarchitecture [10, 22]. Although these reports suggest potential intraoperative benefits, including reduced bleeding and improved visualization, indications, techniques, and reported outcomes remain heterogeneous, precluding definitive conclusions. Our findings—demonstrated in representative cases (see Figs. 1, 2, and 3)—suggest that targeted feeder embolization can safely reduce intraoperative bleeding and facilitate resection, supporting the concept that embolization may benefit selected GBM cases with deep or well-defined pial feeders.

However, the present series should be interpreted within this exploratory context. Our findings are broadly consistent with prior reports describing the potential intraoperative utility of selective feeder embolization in carefully chosen cases, while similarly underscoring the absence of high-level evidence supporting routine application. Accordingly, our contribution is intended to extend the technical and anatomical understanding of this approach rather than to redefine its clinical indications.

Novelty and technical distinction

Several features of our approach warrant emphasis.

First, all procedures were performed via a 3-Fr distal transradial approach, which minimizes access-site morbidity and enables early mobilization [9, 19].

Second, the radiopaque coil or NBCA cast provided a useful intraoperative landmark that aided identification of feeder origins and resection depth (coil in Fig. 1F–G, coil in Fig. 2F, NBCA cast in Fig. 3A–B), complementing earlier case-based observations [10, 22].

Third, the integration of real-time neurological monitoring during super-selective injections ensured that embolization proceeded only when tumor-restricted perfusion was unequivocally confirmed.

The absence of quantified selection criteria represents an inherent limitation of this study. However, establishing universally applicable numerical thresholds for embolization candidacy in glioblastoma is challenging due to substantial intertumoral variability in vascular architecture, feeder distribution, and proximity to eloquent cortex. In this context, rigid quantitative criteria may be less clinically meaningful than detailed angiographic and anatomical judgment.

Accordingly, the heterogeneity in tumor location, feeder anatomy, and embolic materials observed in this series reflects real-world clinical diversity rather than methodological inconsistency. The present study was not designed to isolate the effect of a single anatomical or material-specific variable, but rather to demonstrate the feasibility and safety of selective feeder embolization across a range of carefully selected, anatomically diverse glioblastoma cases.

Benefits of preoperative embolization

In extra-axial tumors such as meningiomas, embolization is known to reduce operative time and blood loss [2, 10]. In the present series, similar intraoperative trends were descriptively observed, although direct comparison was not performed.

We fully agree that early identification and control of tumor feeders during microsurgical resection—sometimes referred to as “surgical embolization”—is a fundamental and highly effective strategy for minimizing blood loss and maintaining a clean operative field in glioma surgery. Meticulous microsurgical technique and early feeder control remain cornerstones of safe glioblastoma resection.

In this context, preoperative embolization should not be viewed as a substitute for surgical hemostatic strategies, but rather as a complementary adjunct in selected cases. Specifically, it may be useful when major feeders are deeply situated or surgically inaccessible early in the procedure, or when complex vascular anatomy limits safe early surgical control. Under such circumstances, preoperative flow reduction may facilitate subsequent microsurgical dissection while preserving established surgical principles.

In the present series, flow reduction along pial feeders was associated with decreased congestion at the tumor surface and improved visualization, facilitating precise dissection of deep vascular structures. These intraoperative features are illustrated in Figs. 1F–I and 2F–I, where the absence of arterial back-bleeding and improved anatomic orientation enabled preservation of venous structures and efficient tumor removal.

Although reduced intraoperative bleeding and improved surgical conditions were consistently observed, the present study was not designed to quantitatively evaluate operative efficiency or invasiveness relative to institutional benchmarks or non-embolized glioblastoma cases. Such comparisons would be highly susceptible to confounding by tumor location, vascular architecture, surgical complexity, and institutional workflow. Operative time and blood loss were therefore reported descriptively to characterize procedural features rather than to demonstrate superiority or reduced invasiveness.

Challenges unique to glioblastoma

Unlike extra-axial hypervascular tumors supplied by ECA branches, GBM receives blood flow from fine-caliber pial or perforating arteries. These vessels often share territories with eloquent cortex, limiting the indications for safe embolization.

Consistent with prior reports [8], accurate delineation of perfusion boundaries during super-selective angiography was essential. Our experience reinforces that embolization should be reserved for tumors with clearly defined feeders and minimal physiological territory—anatomical situations demonstrated in Figs. 1B–D and 2B–E.

Safety considerations

Although modern cerebral angiography has a low complication rate, adherence to strict selection criteria and procedural safeguards remains crucial. Our findings were consistent with contemporary safety benchmarks [7, 12].

Embolization of pial or perforator feeders in glioblastoma differs fundamentally from embolization of extra-axial tumors and carries inherent risks of ischemic injury to eloquent cortex. The absence of ischemic complications in the present series should therefore not be interpreted as evidence that such risks are negligible. Rather, the observed safety profile reflects strict case selection and a deliberately conservative embolization strategy prioritizing functional preservation over angiographic completeness.

These considerations underscore that preoperative embolization for glioblastoma should be reserved for highly selected cases and performed only with meticulous angiographic assessment, real-time neurological monitoring, and a clear understanding of vascular functional territories.

The absence of new ischemic lesions on postoperative MRI across the series (summarized in Figs. 1J and 2J) underscores the safety of selective feeder embolization when applied within well-defined indications.

Histopathology and 5-ALA fluorescence

The potential impact of preoperative feeder embolization on histopathological interpretation and 5-ALA–induced fluorescence remains an important theoretical consideration. Because the present study did not include a contemporaneous non-embolized control group, no definitive conclusions can be drawn regarding differential effects of embolization on fluorescence intensity or tissue characteristics. Accordingly, our observations are intended to provide contextual insight rather than evidence of a causal relationship.

Prior studies indicate that selective embolization preserves diagnostic tissue quality [13] and that 5-ALA fluorescence remains reliable even in reactive or post-treatment tissue [3, 11, 15]. Consistent with these observations, preserved histopathological features and adequate 5-ALA fluorescence were observed in this series; however, these findings should be interpreted as hypothesis-generating. Systematic comparative studies will be required to clarify whether selective embolization influences fluorescence patterns or pathological assessment in glioblastoma surgery.

Timing of embolization

The next-day resection protocol (within 24 h) was chosen to maintain durable devascularization while minimizing the chance of recanalization or inflammatory change. This timing was supported by consistent intraoperative findings across cases (Figs. 1F–I, 2F–I and 3A–B) and aligns with experience reported in meningioma literature [9].

Practicality and reproducibility

Although this is a single-center experience, the technique uses standard endovascular tools—3-Fr sheaths, low-profile microcatheters, detachable coils, and NBCA—available at most neurointerventional centers. The short procedure times and clear intraoperative advantages (demonstrated in Figs. 1, 2, and 3) suggest that preoperative embolization can be incorporated into treatment workflows without introducing significant delays or complexity.

Broader applicability and future perspectives

Selective preoperative embolization is not intended to replace intraoperative functional mapping or advanced visualization strategies, but rather to complement them in anatomically complex or hypervascular cases [5]. With increasing availability of low-profile endovascular systems and multidisciplinary neuro-oncology teams, barriers to adoption are diminishing.

Future prospective investigations will help clarify the clinical benefit, cost-effectiveness, and workflow implications of incorporating embolization into modern glioblastoma surgery, ultimately defining the subset of patients most likely to benefit from this adjunct.

Limitations

This study has several limitations.

First, this study is retrospective, single-center, and single-arm in nature, involving a limited number of patients without a non-embolized control group. These design characteristics preclude definitive causal inference regarding whether the observed reductions in intraoperative bleeding or operative time can be attributed specifically to preoperative embolization, as opposed to other confounding factors such as tumor vascularity, surgical technique, or institutional workflow.

Second, because postoperative MRI was obtained more than 72 h after surgery, the presence of subclinical ischemic changes caused by embolization cannot be completely excluded. Although no intraoperative findings suggestive of ischemic tissue were observed and postoperative imaging demonstrated no infarction beyond the planned resection margins, the absence of immediate post-embolization imaging represents a limitation of this study. In addition, the small sample size limits the ability to detect rare but potentially devastating embolization-related adverse events.

Third, independent of the imaging-related limitation described above, the lack of a contemporaneous control group and the heterogeneity in tumor location, vascularity, and prior treatments introduce potential selection and performance biases, particularly because all procedures were performed by the same surgical and endovascular team. In addition, the absence of quantified embolization criteria and the heterogeneity in tumor characteristics and embolization techniques may limit reproducibility and complicate interpretation of the findings.

Fourth, operative metrics such as blood loss and microscope-assisted operative time may have been influenced by institutional workflow and surgeon experience rather than embolization alone.

Fifth, long-term oncological and functional outcomes, including progression-free survival, overall survival, and detailed postoperative neurological function, were not evaluated in this study. Accordingly, the present analysis does not allow assessment of the broader oncological or functional impact of preoperative embolization beyond its intraoperative role. The primary focus of this study was technical feasibility, procedural safety, and intraoperative utility; evaluation of survival benefit or long-term neurological outcomes was intentionally beyond its scope.

Prospective studies incorporating standardized survival and neurological endpoints will be required to determine whether the intraoperative advantages observed with selective feeder embolization translate into meaningful long-term clinical benefit.

Future prospective, multi-center studies with standardized endpoints will be essential to clarify patient selection criteria, validate the hemostatic and workflow benefits, and determine the broader clinical role of preoperative feeder embolization in glioblastoma surgery.

Conclusion

Preoperative embolization of glioblastoma-feeding arteries, performed through a minimally invasive distal radial approach under local anesthesia, was associated with reduced intraoperative bleeding and improved anatomical orientation, facilitating safer and more efficient resections in appropriately selected cases. Although limited by sample size and the absence of a direct control group, the consistent technical success and lack of ischemic complications support the procedural safety and feasibility of this approach, particularly for hypervascular or anatomically complex tumors that challenge conventional microsurgery. While not intended for routine use, selective feeder embolization may serve as a valuable adjunct to mapping-guided resection, and further multicenter prospective studies are needed to define optimal patient selection, quantify hemostatic and workflow benefits, and clarify its broader role within modern glioblastoma surgery.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

GBM

Glioblastoma

DSA

Digital subtraction angiography

dTRA

Distal transradial approach

NBCA

N-butyl-2-cyanoacrylate

ETA

Early temporal artery

AIFA

Anterior internal frontal artery

MIFA

Middle internal frontal artery

ICA

Internal carotid artery

MCA

Middle cerebral artery

SMCV

Superficial middle cerebral vein

IAP

Intracarotid amobarbital procedure

MRI

Magnetic resonance imaging

CT

Computed tomography

Author Contribution

Masashi Uchida: Conceptualization, Investigation, Writing – original draft. Hidemichi Ito: Investigation, Writing – review & editing. Yuichiro Kushiro: Investigation. Gaku Hidaka: Investigation. Sora Yazaki: Investigation. Yasuyuki Yoshida: Investigation., Hiroshi Takasuna: Supervision. Takashi Matsumori: investigation. Ichiro Takumi: Supervision. Hidetaka Onodera: Supervision. Toshihiro Ueda: Supervision, Investigation. Hidetoshi Murata: Conceptualization, Supervision, Writing – review & editing. All authors read and approved the final manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data Availability

No datasets were generated or analysed during the current study.

Declarations

Ethical approval

All procedures performed in this study involving human participants were conducted in accordance with the ethical standards of the institutional research committee (IRB No. 7216) and with the 1964 Helsinki Declaration and its later amendments. Written informed consent was obtained from all patients included in this study.

Informed consent

This study was approved by the Institutional Review Board of St. Marianna University School of Medicine (January 14, 2025; No. 7216). Written informed consent was obtained from all subjects (or their legal guardians).

Declaration of generative AI use

The authors did not use generative AI or AI-assisted technologies in the preparation of this manuscript.

Competing interest

The authors declare no competing interests.