Abstract
Embolization of intracranial tumors is predominantly performed in Japan, primarily before neurosurgical resection. The Japanese Registry of NeuroEndovascular Therapy (JR-NET) Study Group, established in 2005, aims to clarify the factors influencing the outcomes of neuroendovascular treatment. Japanese Registry of NeuroEndovascular Therapy 4 is a nationwide, multicenter retrospective observational study that evaluates real-world data on intracranial tumor embolization in Japan. Japanese Registry of NeuroEndovascular Therapy 4 is based on data collected from 166 neurosurgical centers in Japan between January 2015 and December 2019. Of 63,230 patients, 2,664 (4.2%) with intracranial tumors underwent embolization. The primary endpoint was the proportion of patients with a modified Rankin scale (mRS) score of 0-2 at 30 days post-procedure. Secondary endpoints included procedure-related complications. Among the 2,664 patients, 61 records lacked sufficient data, leaving 2,603 patients (1,612 females, median age: 61 years [interquartile range 51-71]). The proportion of patients with mRS scores ≤2 at 30 days after the procedure was 86.9%. The overall incidence of procedure-related complications was 4.8%, with 1.8% hemorrhagic, 2.0% ischemic, and 1.0% classified as other complications. In the multivariate analysis, general anesthesia and embolization of vessels other than the external carotid artery were identified as risk factors for the development of complications. Meningioma cases had a complication rate of 4.3%, with major complications occurring in 3.5%. Hemangioblastoma cases had a 14.9% complication rate, with major complications at 9.9%. Japanese Registry of NeuroEndovascular Therapy 4 provides comprehensive real-world data on intracranial tumor embolization in Japan, identifying risk factors to inform and improve the safe practice of intracranial tumor embolization in neuroendovascular therapy.
Keywords: complication, intracranial tumor, embolization, Japanese registry
Introduction
Neuroendovascular treatment for intracranial tumors, especially preoperative embolization for meningioma, is reportedly a standard procedure to reduce intraoperative blood loss and facilitate resection.1) However, many contemporary reports are single-center retrospective analyses with potential selection bias.2)
The multicenter Japanese Registry of NeuroEndovascular Therapy (JR-NET) Study Group was formed in 2005 and clarified the factors affecting the outcomes of neuroendovascular treatment.3) JR-NET4 was the fourth study in a series of retrospective, nationwide, multicenter, observational studies on neuroendovascular treatment in Japan. The JR-NET2 and JR-NET3 studies assessed the real-world outcomes of intracranial tumor embolization in Japan using a large dataset.4,5) In JR-NET2 and JR-NET3, the incidence rates of complications in patients who underwent intracranial tumor embolization were 1.5% (15 of 1,012) and 3.7% (57 of 1,337). Upon multivariate analysis, embolization for tumors other than meningioma was the only significant risk factor for complications in JR-NET2 (odds ratio [OR], 3.88; 95% confidence interval [CI], 1.13-12.10; p = 0.032) while embolization for feeders other than the external carotid artery (ECA; OR, 3.56; 95% CI, 2.03-6.25; p < 0.001) and the use of liquid materials (OR, 2.65; 95% CI, 1.50-4.68; p < 0.001) were risk factors for the development of complications in JR-NET3. Since the publication of the JR-NET3 study, intracranial tumor embolization has become more popular, and the number of cases has increased. We therefore considered that an increase in the number of enrolled cases would provide a more accurate analysis of intracranial tumor embolization.
In this study, we compared JR-NET3 and JR-NET4 to clarify the real-world data on embolization for intracranial tumors in Japanese patients. Furthermore, for the first time in a JR-NET study, we analyzed tumor-specific embolization in JR-NET4.
Materials and Methods
Patients and methods
Patients were derived from the JR-NET4 database. This was a retrospective, observational study using data from 166 neurosurgical centers in Japan, collected from January 2015 to December 2019. In total, 63,230 patients were enrolled in JR-NET4; of these, 2,664 (4.2%) with intracranial tumors underwent embolization. The primary endpoint was the proportion of patients with a modified Rankin scale (mRS) score of 0-2 (functional independence) at 30 days post-procedure. The secondary endpoint was the occurrence of complications related to the procedures. The outcomes of complications were classified as minor and major, as in the previous JR-NET3; minor complications included no or transient symptoms with complete remission within 30 days, and major complications included a >1-point decrease in mRS score.
Age, sex, and preoperative mRS score were recorded as background variables. The types of tumors, target vessels for embolization, embolic materials, and embolization results were evaluated. The target vessels were categorized as feeders from the internal carotid artery (ICA), ECA, and vertebrobasilar artery (VBA). The embolic materials used were coils, liquid materials such as n-butyl cyanoacrylate and ONYX, particle materials such as polyvinyl alcohol, and combinations of these. The results of each embolization were indicated by the degree of devascularization in each tumor, which was graded as total, subtotal, partial, or unchanged. This grading was based on the operator's visual interpretation of angiographic findings, and no standardized quantitative definition was provided in the registry protocol.
Complications were defined as any neurological deficit or death that occurred during or after embolization. Moreover, radiographical abnormalities after embolization, such as ischemic or hemorrhagic changes, were classified as complications even if the patients were asymptomatic. In patients with complications, the type of complication, timing and duration of its occurrence, treatments used, and outcomes of the complication were recorded. Morbidity included complications with continuous symptoms within 30 days after the procedure, and mortality included an mRS score of 6 within 30 days. Details and outcomes of direct surgery or radiation therapy for intracranial tumors were not collected in this study. In this study, JR-NET2, 3 and 4 were first compared, and further tumor-specific analyses were performed for JR-NET4.
Statistical analysis
All statistical analyses were performed using JMP 10 software (SAS Institute Inc.). Descriptive statistics are expressed as median (interquartile range [IQR]). The univariate associations between each potential risk factor and the occurrence of complications and death (mRS score of 6 within 30 days post-procedure) were assessed using Fisher's exact test for cell sizes <10 and the Wilcoxon rank sum test for others. Potential risk factors with p < 0.05 on univariate analysis were included as confounders in the multivariate logistic regression model to determine whether or not they were risk factors for complications and death. The results are presented as ORs with 95% CIs. p-Values < 0.05 were considered indicative of statistical significance. Furthermore, variables associated with p < 0.05 in the univariate analysis and those identified as risk factors for complications in previous reports (JR-NET2 or JR-NET3) were selected for multivariate analysis.4,5)
Results
Of 2,664 patients registered with intracranial tumor embolization in JR-NET4, the records for 61 patients lacked sufficient information for this evaluation. Thus, 2,603 patients (1,612 females, median age: 61 years [IQR 51-71]) were included in this study after excluding one patient with insufficient information. Of the tumors enrolled, 2,232 (85.7%) were meningiomas, 141 (5.4%) were hemangioblastomas, 21 (0.8%) were gliomas, and 209 (8.0%) were other tumors. Procedures for 473 (18.2%) patients were performed under general anesthesia, and for 2,130 (81.8%) patients under local anesthesia. The patient characteristics and procedures performed are summarized in Table 1. In JR-NET4, target vessels for procedures were ECA in 2128 (81.8%), ICA in 127 (4.9%), VBA in 188 (7.2%), ECA + ICA in 117 (4.5%), ECA + VBA in 25 (1.0%), ICA + VBA in 6 (0.2%), ECA + ICA + VBA in 3 (0.1%), and other vessels in 9 (0.3%) cases. The total number of embolic materials used was 1,508 for coils (57.9%), 1416 for particles (54.4%), and 957 for liquid agents (36.8%).
Table 1.
Patient Characteristics in JR-NET4
| Patients (n = 2,603) | ||
|---|---|---|
| Age, years (median, IQR) | 61 (51-71) | |
| Sex, female | 1,612 | 61.9% |
| Preoperative mRS score | ||
| 0 | 1,542 | 59.2% |
| 1 | 537 | 20.6% |
| 2 | 307 | 11.8% |
| 3 | 139 | 5.3% |
| 4 | 71 | 2.7% |
| 5 | 7 | 0.3% |
| Type of tumors | ||
| Meningioma | 2,232 | 85.7% |
| Hemangioblastoma | 141 | 5.4% |
| Glioma | 21 | 0.8% |
| Others | 209 | 8.0% |
| Anesthesia | ||
| General | 473 | 18.2% |
| Local | 2,130 | 81.8% |
| Scheduled intervention | ||
| Yes | 2,569 | 98.7% |
| No | 34 | 1.3% |
| Target vessels | ||
| ECA | 2,128 | 81.8% |
| ICA | 127 | 4.9% |
| VBA | 188 | 7.2% |
| ECA + ICA | 117 | 4.5% |
| ECA + VBA | 25 | 1.0% |
| ICA + VBA | 6 | 0.2% |
| ECA + ICA + VBA | 3 | 0.1% |
| Others | 9 | 0.3% |
| Embolic materials | ||
| Coil | 1,508 | 57.9% |
| Particle | 1,416 | 54.4% |
| Liquid | 957 | 36.8% |
| Results of embolization | ||
| Total | 525 | 20.2% |
| Subtotal | 1,110 | 42.6% |
| Partial | 944 | 36.3% |
| Unchanged | 16 | 0.6% |
Data presented as frequency (percentage) unless indicated otherwise.
ECA: external carotid artery; ICA: internal carotid artery; IQR: interquartile range; JR-NET4: Japanese Registry of NeuroEndovascular Therapy 4; mRS: modified Rankin scale; VBA: vertebrobasilar artery
Table 2 presents a comparison of patient characteristics and incidence of endpoints between JR-NET2, JR-NET3, and JR-NET4. Table 3 compares the complications of each JR-NET study. Among 2,603 patients in JR-NET4, 125 (4.8%) developed procedural complications (secondary endpoint). Of the 125 cases with complications, 46 (1.8%) were hemorrhagic, 53 (2.0%) were ischemic, and 26 (1.0%) involved other types of complications. The primary endpoint (mRS score 0-2 at 30 days post-procedure) was observed in 2,263 patients (86.9％). In total, 58 (2.2%) patients with complications were asymptomatic, 16 (0.6%) developed temporary symptoms, and 50 (1.9%) developed permanent symptoms. Of the 50 patients with permanent symptoms, the mRS score at 30 days declined by 1 point in 26 (1.0%) patients and by more than 2 points in 24 (0.9%) patients.
Table 2.
Comparison of Patient Characteristics and Incidence of Endpoints between JR-NET2, JR-NET3, and JR-NET4
| JR-NET2 (n = 1,012) |
JR-NET3 (n = 1,544) |
JR-NET4 (n = 2,603) |
||||
|---|---|---|---|---|---|---|
| Age, years (median, IQR) | 60 | 50-68 | 63 | 52-70 | 61 | 51-71 |
| Female | 637 | 62.9% | 937 | 60.7% | 1,612 | 61.9% |
| Preoperative mRS score | ||||||
| 0 | 696 | 68.8% | 1,034 | 67.0% | 1,542 | 59.2% |
| 1 | 173 | 17.1% | 271 | 17.5% | 537 | 20.6% |
| 2 | 77 | 7.6% | 137 | 8.9% | 307 | 11.8% |
| 3 | 26 | 2.6% | 58 | 3.7% | 139 | 5.3% |
| 4 | 25 | 2.4% | 34 | 2.2% | 71 | 2.7% |
| 5 | 3 | 0.3% | 6 | 0.4% | 7 | 0.3% |
| Type of tumors | ||||||
| Meningioma | 898 | 88.7% | 1,337 | 86.6% | 2,232 | 85.7% |
| Hemangioblastoma | 25 | 2.5% | 72 | 4.7% | 141 | 5.4% |
| Glioma | 6 | 0.6% | 10 | 0.6% | 21 | 0.8% |
| Others | 78 | 7.7% | 121 | 7.8% | 209 | 8.0% |
| Anesthesia | ||||||
| General | 42 | 4.2% | 203 | 13.1% | 473 | 18.2% |
| Local | 968 | 94.4% | 1,338 | 86.7% | 2,130 | 81.8% |
| Scheduled intervention | ||||||
| Yes | 966 | 98.4% | 1,520 | 98.4% | 2,569 | 98.7% |
| No | 16 | 1.6% | 21 | 1.4% | 34 | 1.3% |
| Target vessels | ||||||
| ECA only | 908 | 89.7% | 1,294 | 83.8% | 2,128 | 81.8% |
| Others than ECA | 92 | 9.1% | 233 | 15.1% | 475 | 18.2% |
| Embolic materials | ||||||
| Coil | 607 | 60.0% | 827 | 53.6% | 1,508 | 57.9% |
| Particle | 596 | 58.9% | 632 | 40.9% | 1,416 | 54.4% |
| Liquid | 214 | 21.1% | 627 | 40.6% | 957 | 36.8% |
| Results of embolization | ||||||
| Total | 344 | 34.0% | 385 | 24.9% | 525 | 20.2% |
| Subtotal | 414 | 40.9% | 627 | 40.6% | 1,110 | 42.6% |
| Partial | 238 | 23.5% | 492 | 31.9% | 944 | 36.3% |
| Unchanged | 7 | 0.7% | 16 | 1.0% | 16 | 0.6% |
| Type of complications | ||||||
| Total | 15 | 1.5% | 57 | 3.7% | 125 | 4.8% |
| Hemorrhage | 4 | 0.4% | 7 | 0.5% | 46 | 1.8% |
| Ischemia | 6 | 0.6% | 32 | 2.2% | 53 | 2.0% |
| Others | 5 | 0.5% | 18 | 1.2% | 26 | 1.0% |
| Outcome of complications | ||||||
| Minor | 13 | 1.3% | 38 | 2.5% | 75 | 2.9% |
| Major | 2 | 0.2% | 17 | 1.1% | 50 | 1.9% |
| Data not acquired | 0 | 0.0% | 2 | 0.1% | 0 | 0.0% |
Data presented as frequency (percentage) unless indicated otherwise.
ECA: external carotid artery; IQR: interquartile range; JR-NET: Japanese Registry of NeuroEndovascular Therapy; mRS: modified Rankin scale
Table 3.
Patient Characteristics and Type of Tumor in JR-NET4
| Meningioma (n = 2,232) |
Hemangioblastoma (n = 141) |
Glioma (n = 21) |
||||
|---|---|---|---|---|---|---|
| Age, years (median, IQR) | 62 (52-72) | 54 (41-66) | 55 (43-71) | |||
| Sex, female | 1,437 | 64.4% | 58 | 41.1% | 10 | 47.6% |
| Preoperative mRS score | ||||||
| 0 | 1,341 | 60.1% | 72 | 51.1% | 8 | 38.1% |
| 1 | 454 | 20.3% | 37 | 26.2% | 4 | 19.0% |
| 2 | 258 | 11.6% | 16 | 11.3% | 3 | 14.3% |
| 3 | 113 | 5.1% | 10 | 7.1% | 4 | 19.0% |
| 4 | 59 | 2.6% | 6 | 4.3% | 2 | 9.5% |
| 5 | 7 | 0.3% | 0 | 0.0% | 0 | 0.0% |
| Anesthesia | ||||||
| General | 352 | 15.8% | 68 | 48.2% | 5 | 23.8% |
| Local | 1,880 | 84.2% | 73 | 51.8% | 16 | 76.2% |
| Scheduled intervention | ||||||
| Yes | 2,207 | 98.9% | 135 | 95.7% | 21 | 100.0% |
| No | 25 | 1.1% | 6 | 4.3% | 0 | 0.0% |
| Target vessels | ||||||
| ECA | 1,953 | 87.5% | 18 | 12.8% | 6 | 28.6% |
| ICA | 95 | 4.3% | 1 | 0.7% | 8 | 38.1% |
| VBA | 44 | 2.0% | 116 | 82.3% | 6 | 28.6% |
| ECA + ICA | 109 | 4.9% | 1 | 0.7% | 1 | 4.8% |
| ECA + VBA | 20 | 0.9% | 4 | 2.8% | 0 | 0.0% |
| ICA + VBA | 4 | 0.2% | 1 | 0.7% | 0 | 0.0% |
| ECA + ICA + VBA | 3 | 0.1% | 0 | 0.0% | 0 | 0.0% |
| Others | 4 | 0.2% | 0 | 0.0% | 0 | 0.0% |
| Embolic materials | ||||||
| Coil | 1,336 | 59.9% | 46 | 32.6% | 12 | 57.1% |
| Particle | 1,304 | 58.4% | 17 | 12.1% | 3 | 14.3% |
| Liquid | 746 | 33.4% | 110 | 78.0% | 14 | 66.7% |
| Results of embolization | ||||||
| Total | 475 | 21.3% | 19 | 13.5% | 2 | 9.5% |
| Subtotal | 953 | 42.7% | 64 | 45.4% | 8 | 38.1% |
| Partial | 789 | 35.3% | 56 | 39.7% | 11 | 52.4% |
| Unchanged | 12 | 0.5% | 1 | 0.7% | 0 | 0.0% |
| Complications | 2,232 | 141 | 21 | |||
| Total | 95 | 4.3% | 21 | 14.9% | 2 | 9.6% |
| Hemorrhage | 32 | 1.4% | 10 | 7.1% | 1 | 4.8% |
| Ischemia | 41 | 1.8% | 9 | 6.4% | 1 | 4.8% |
| Others | 22 | 1.0% | 2 | 1.4% | 0 | 0.0% |
| Outcome of complications | ||||||
| Minor | 60 | 2.7% | 10 | 7.1% | 1 | 4.8% |
| Major | 35 | 1.6% | 11 | 7.8% | 1 | 4.8% |
| Data not acquired | 0 | 0.0% | 0 | 0.0% | 0 | 0.0% |
Data presented as frequency (percentage) unless indicated otherwise.
ECA: external carotid artery; ICA: internal carotid artery; IQR: interquartile range; JR-NET4: Japanese Registry of NeuroEndovascular Therapy 4; mRS: modified Rankin scale; VBA: vertebrobasilar artery
Patient characteristics disaggregated by tumor type are summarized in Table 4. Target vessels for meningioma procedures were the ECA in 1,953 (87.5%), ICA in 95 (4.3%), VBA in 44 (2.0%), ECA + ICA in 109 (4.9%), ECA + VBA in 20 (0.9%), ICA + VBA in 4 (0.2%), ECA + ICA + VBA in 3 (0.1%), and others vessels in 4 (0.2%) cases. Target vessels for hemangioblastoma procedures were the ECA in 18 cases (12.8%), ICA in 1 (0.7%), VBA in 116 (82.3%), ECA + ICA in 1 (0.7%), ECA + VBA in 4 (2.8%), and ICA + VBA in 1 (0.7%) case. For glioma procedures, target vessels were the ECA in 6 (28.6%), ICA in 8 (38.1%), VBA in 6 (28.6%), and ECA + ICA in 1 (4.8%) case. The embolic materials used for meningioma were 1,336 for coils (59.9%), 1,416 for particles (63.4%), and 957 for liquid (42.9%). For hemangioblastoma, embolic materials included 46 for coils (32.6%), 17 for particles (12.1%), and 110 for liquid agents (78.0%). For gliomas, the embolic materials used were 12 for coils (57.1%), 3 for particles (14.3%), and 14 for liquid agents (66.7%).
Table 4.
Univariate and Multivariate Analysis: Risk Factors Related to the Development of Complications in JR-NET4
| Factors | Univariate analysis | Multivariate analysis | ||||||
|---|---|---|---|---|---|---|---|---|
| Occurrence of complications | p-Value | OR | 95% CI | p-Value | ||||
| Yes | No | |||||||
| Number of patients | 125 | 4.8% | 2,478 | 95.2% | ||||
| Age (median, IQR) | 60 (50-70) | 61 (51-71) | 0.74 | |||||
| Female sex | 70 | 2.7% | 1,542 | 59.2% | 0.19 | |||
| Preoperative mRS 3-5 | 14 | 0.5% | 203 | 7.8% | 0.24 | |||
| Other than meningioma | 30 | 1.2% | 341 | 13.1% | 0.0025 | 1.13 | 0.70-1.84 | 0.6 |
| Scheduled procedure | 123 | 4.7% | 2,446 | 94.0% | 0.68 | |||
| General anesthesia | 42 | 1.6% | 431 | 16.6% | <0.0001 | 1.86 | 1.22-2.84 | 0.0037 |
| Other than ECA | 49 | 1.9% | 286 | 11.0% | <0.0001 | 2.58 | 1.67-3.98 | <0.0001 |
| Coil | 60 | 2.3% | 1,448 | 55.6% | 0.026 | |||
| Particle | 57 | 2.2% | 1,359 | 52.2% | 0.053 | |||
| Liquid | 58 | 2.2% | 899 | 34.5% | 0.028 | 0.97 | 0.65-1.44 | 0.87 |
| Complete embolization | 18 | 0.7% | 507 | 19.5% | 0.11 | |||
Data presented as frequency unless indicated otherwise.
CI: confidence interval; ECA: external carotid artery; IQR: interquartile range; JR-NET4: Japanese Registry of NeuroEndovascular Therapy 4; mRS: modified Rankin scale; OR: odds ratio
In total, 125 patients experienced treatment-related complications. Among patients with meningioma, complications were observed in 95 patients (4.3%), with 35 (1.6%) developing major complications, including hemorrhagic complications in 32 patients (1.4%), ischemic complications in 41 (1.8%), and other complications in 22 (1.0%). Among patients with hemangioblastoma, 21 cases (14.9%) involved treatment-related complications, with 11 cases (7.8%) resulting in major complications. These included hemorrhagic complications in 10 cases (7.1%), ischemic in 9 (6.4%), and other complications in 2 (1.4%). Among patients with glioma, there were 2 cases (9.6%) of treatment-related complications, with 1 case (4.8%) resulting in major complications, including hemorrhagic and ischemic complications in 1 case (4.8%) each.
Risk factors for complications
The results of univariate analysis and multivariate analysis are shown in Table 4. In patients undergoing tumor embolization, tumor type other than meningioma, general anesthesia, target vessels other than ECA, and the use of coil and liquid material were associated with the development of complications. Upon multivariate analysis, only general anesthesia (OR, 1.86; 95% CI, 1.22-2.84; p = 0.037) and target vessels other than ECA (OR, 2.58; 95% CI, 1.67-3.98; p < 0.0001) were significant risk factors for complications. The results of univariate and multivariate analyses by tumor type are shown in Table 5. In meningioma, target vessels other than the ECA were significantly associated with the development of complications (OR, 2.42; 95% CI, 1.47-3.99; p = 0.0005). In hemangioblastoma, age and general anesthesia were significantly associated with the development of complications. Upon multivariate analysis, only general anesthesia (OR, 10.63; 95% CI, 2.67-42.29; p = 0.0008) was a significant risk factor for complications. Glioma was not analyzed due to the small number of cases.
Table 5.
Univariate and Multivariate Analysis: Risk Factors Related to the Development of Complications for Meningioma, Hemangioblastoma, and Glioma
| Meningioma | ||||||||
|---|---|---|---|---|---|---|---|---|
| Factors | Univariate analysis | Multivariate analysis | ||||||
| Occurrence of complications | p-Value | OR | 95%CI | p-Value | ||||
| Yes | No | |||||||
| Number of patients | 95 | 4.3% | 2,137 | 95.7% | ||||
| Age (median, IQR) | 61 (50-71) | 62 (52-72) | 0.6 | |||||
| Female sex | 57 | 2.6% | 1,380 | 61.8% | 0.38 | |||
| Preoperative mRS 3-5 | 10 | 0.4% | 169 | 7.6% | 0.33 | |||
| Scheduled procedure | 95 | 4.3% | 2,112 | 94.6% | 0.60 | |||
| General anesthesia | 22 | 1.0% | 330 | 14.8% | 0.06 | 1.51 | 0.90-2.54 | 0.12 |
| Other than ECA | 24 | 1.1% | 129 | 5.8% | <0.001 | 2.42 | 1.47-3.99 | 0.0005 |
| Coil | 49 | 2.2% | 1,287 | 57.7% | 0.11 | |||
| Particle | 55 | 2.5% | 1,249 | 56.0% | 0.92 | |||
| Liquid | 35 | 1.6% | 711 | 31.9% | 0.51 | 0.9 | 0.57-1.42 | 0.65 |
| Complete embolization | 16 | 0.7% | 459 | 20.6% | 0.31 | |||
| Hemangioblastoma | ||||||||
|---|---|---|---|---|---|---|---|---|
| Factors | Univariate analysis | Multivariate analysis | ||||||
| Occurrence of complications | p-Value | OR | 95% CI | p-Value | ||||
| Yes | No | |||||||
| Number of patients | 21 | 14.9% | 120 | 85.1% | ||||
| Age (median, IQR) | 60 (54-70) | 53 (40-66) | 0.032 | |||||
| Female sex | 8 | 5.7% | 50 | 35.5% | 0.81 | |||
| Preoperative mRS 3-5 | 4 | 2.8% | 12 | 8.5% | 0.26 | |||
| Scheduled procedure | 19 | 13.5% | 116 | 82.3% | 0.22 | |||
| General anesthesia | 18 | 12.8% | 50 | 35.5% | <0.001 | 10.63 | 2.67-42.29 | 0.0008 |
| Other than ECA | 19 | 13.5% | 104 | 73.8% | >0.99 | 0.67 | 0.11-3.99 | 0.66 |
| Coil | 5 | 3.5% | 41 | 29.1% | 0.45 | |||
| Particle | 1 | 0.7% | 16 | 11.3% | 0.47 | |||
| Liquid | 1 | 0.7% | 94 | 66.7% | 0.78 | 0.5 | 0.14-1.74 | 0.27 |
| Complete embolization | 1 | 0.7% | 18 | 12.8% | 0.31 | |||
Data presented as frequency unless indicated otherwise.
CI: confidence interval; ECA: external carotid artery; IQR: interquartile range; mRS: modified Rankin scale; OR: odds ratio
Discussion
Recent trends in preoperative embolization for intracranial tumors
Preoperative embolization for intracranial tumors is an adjunctive procedure before craniotomy, with the following primary objectives: (1) safe tumor resection and improvement in resection extent by inducing tumor necrosis; (2) reduction of intraoperative bleeding, and (3) shortening of surgical duration.1,6,7) Several studies have demonstrated the utility of preoperative embolization for highly vascular tumors, such as meningiomas.8,9) However, some studies could not demonstrate the effectiveness of preoperative embolization. In a randomized controlled trial by Iacobucci et al.,10) preoperative embolization significantly reduced the surgical duration; however, its benefits regarding complications and bleeding were not justified. Moreover, in Przybylowski et al.'s2) study, preoperative embolization for World Health Organization grade I intracranial meningiomas did not improve the surgical outcomes. Previous retrospective cohort studies have focused on intraoperative bleeding, operative time, and removal rate. However, drawing definitive conclusions is challenging due to considerable variability in factors such as tumor size, location, and operative-related variables.6,11) Akimoto et al.12) conducted a propensity-matched analysis to investigate the effectiveness of preoperative embolization for meningioma resection and found that although preoperative embolization did not influence the tumor resection rate or patient outcomes, the embolization group showed significantly reduced intraoperative bleeding, shortened surgical duration, and longer time to meningioma recurrence. Future studies should focus on intraoperative blood loss, operative time, and removal rate, as in this report, as well as on various other analyses that may demonstrate the efficacy of preoperative embolization.
Factors related to the development of complications
The number of participants in the JR-NET studies on preoperative embolization for brain tumors has significantly increased over time (1,012 in JR-NET2, 1,544 in JR-NET3, and 2,603 in JR-NET4), indicating a substantial expansion in study scale. The JR-NET study is the most extensive report on brain tumor embolization to date.
The secondary endpoint in the JR-NET studies―the procedure-related complication rate (major complication rate)―was 1.5% (0.2%) in JR-NET2, 3.7% (1.1%) in JR-NET3, and 4.8% (1.9%) in JR-NET4.4,5) Furthermore, the JR-NET4 study examined the risk factors for treatment complications as the primary evaluation criteria, consistent with previous JR-NET studies. In JR-NET4, 2 significant risk factors for treatment complications were identified: embolization of vessels other than the ECA and general anesthesia. In this study, the treating physician was recorded; however, the experience of the actual operator performing the procedure was not directly documented. Previous JR-NET studies have suggested that treatment by board-certified specialists or consultation with experts may contribute to a lower rate of complications. Therefore, it is conceivable that the accumulation of operator experience and improvement in technical proficiency may have influenced the overall outcomes in JR-NET4 as well, warranting further investigation.
The risk factors for complications identified herein were consistent with previous JR-NET studies concerning embolization of vessels other than the ECA. Specifically, the highest OR of 2.58 was observed for these target vessels. Intracranial embolization procedures are generally more technically challenging than extracranial embolization.13) Furthermore, practitioners must have in-depth knowledge of intracranial vascular anatomy in addition to technical factors.14) Therefore, in cases requiring intracranial artery embolization, carefully weighing its risks and benefits is imperative.15) Although embolization of the ECA is relatively safe, potential anastomoses with intracranial vessels and neurovascular structures have been reported, indicating that it is not entirely risk-free.16) Therefore, practitioners must be well-versed in these aspects to prevent complications while embolizing these vessels.
Regarding general anesthesia as a risk factor, general anesthesia itself does not pose a threat but potentially reflects the fact that cases undergoing aggressive embolization procedures are more likely to require general anesthesia. In addition, Przybylowski et al.2) reported that the risk of complications due to embolization increases with more aggressive tumor devascularization. Recent improvements in devices have enabled access to more peripheral vessels. Therefore, while aggressive tumor devascularization is now possible, recognizing the associated increasing risks is crucial.
In JR-NET4, the use of liquid embolic materials, identified as a risk factor for complications in JR-NET2 and 3,17) was a risk factor in univariate analysis but not in multivariate analysis. This may be attributable to increased awareness among operators regarding the results of the previous JR-NET studies, resulting in more cautious use of liquid embolic materials. In summary, although a statistical comparison between the JR-NET3 and JR-NET4 trials is not presented, the percentage of cases requiring general anesthesia significantly increased from 13.1% in JR-NET3 to 18.2% in JR-NET4; the rate of target vessels other than the ECA increased significantly from 15.1% in JR-NET3 to 18.2% in JR-NET4, and the proportion of liquid embolic material use significantly decreased from 40.6% in JR-NET3 to 36.8% in JR-NET4. This is likely attributable to an increase in aggressive embolization alongside a more cautious use of liquid embolic materials. The lack of statistical significance in JR-NET4 may be attributable to increased operator awareness and caution in selecting embolic materials, especially following the risk profile identified in previous JR-NET studies. Moreover, advancements in catheter devices and techniques could have contributed to safer delivery and control of liquid embolic agents.
Tumor-specific risk factors associated with embolization procedures
This study analyzed tumor-specific embolization-related factors for meningiomas, hemangioblastoma, and gliomas, which were not examined in previous JR-NET studies. First, no risk factors for procedure-related complications were identified for glioma. Due to the small number of enrolled glioma cases, larger studies are required to obtain more definitive evidence.
The overall complication rate of preoperative embolization for meningiomas was 4.3%, with major complications accounting for 1.6%. In recent studies, the complication rates for preoperative embolization ranged from 5.8% to 8.5%,1,9) suggesting that preoperative embolization for meningiomas is performed relatively safely in Japan. In JR-NET4, the only risk factor for complications of preoperative embolization for meningiomas was embolization of vessels other than ECA, and safe treatment was achieved regardless of the embolization material used. The risks of embolization of vessels other than the ECA have been reported, particularly for embolization of skull base meningiomas. According to Ilyas et al.,18) due to the complex and variable vascular supply to skull base meningiomas, aggressive embolization may increase the risk of severe complications. Thus, meningioma lesions with potential embolization from sources other than the ECA, such as skull base meningiomas, are particularly at risk and should be treated with a thorough understanding of the vascular anatomy.14,16) In addition to the conventionally reported benefits of embolization for meningiomas, it may prolong the recurrence-free period of meningiomas.1,12) Despite its various effects, preoperative embolization for meningiomas is an adjunctive treatment to craniotomy and resection. Therefore, when deciding on a treatment method, considering the benefits and drawbacks of embolization is desirable. As this study did not include detailed surgical outcome data, interpretation of postprocedural mRS should be limited to procedural safety and early postoperative functional status, rather than long-term treatment efficacy. Although postprocedural mRS can be influenced by the patient's preoperative condition and the outcome of subsequent surgical treatment, we selected it as a pragmatic and consistent outcome measure to evaluate the short-term functional impact of embolization across a nationwide registry. This choice was based on the precedent set by previous JR-NET studies and the feasibility of uniformly collecting mRS data at 30 days.
Complications of preoperative embolization for hemangioblastoma include medullary infarction, cerebellar infarction, increased intracranial pressure due to postembolization edema, and hemorrhage.7) Although a meta-analysis evaluated the outcomes of embolization for hemangioblastoma,19) no previous study has reported >140 cases, making this the most significant case series following JR-NET3. In JR-NET4, the overall complication rate of embolization for hemangioblastoma was 14.9%, with major complications accounting for 7.8%. A previous meta-analysis reported a major complication rate of 11.7% for embolization, indicating similar outcomes.7) In JR-NET4, the only risk factor for complications in preoperative embolization was general anesthesia, with an OR of 10.68, significantly higher than meningiomas. Liquid embolic agents were used in 78.0% of cases; however, the use of liquid embolic agents was not identified as a risk factor. Furthermore, 82.3% of the target vessels for hemangioblastoma embolization were in the VBA, but the target vessel was not a risk factor. The fact that general anesthesia was the only risk factor suggests that embolization for hemangioblastoma is an aggressive tumor devascularization procedure, regardless of the target vessels or embolic materials used. This elevated value may reflect selection bias, where general anesthesia was preferentially used in more complex or high-risk cases requiring aggressive embolization. Therefore, this association should be interpreted with caution. This high complication rate may be partially explained by the fact that embolization procedures for hemangioblastomas are often technically demanding and require general anesthesia, reflecting the aggressiveness and complexity of these cases. More peripheral vessel embolization has become possible with the development of devices such as small microcatheters and intermediate catheters. This allows for embolization of only the peripheral feeding vessels of hemangioblastoma, a more aggressive tumor devascularization approach requiring caution. Further research is required to clarify which hemangioblastomas should and should not be embolized. Following this, validation of the presented findings with a randomized controlled trial may be considered.
Limitations
First, this was a retrospective registry study, which is subject to inherent selection bias and the risk of missing or incomplete data. Second, data on the size and location of tumors were not included in the analysis. Tumor location and size may be risk factors for embolization complications, particularly in meningiomas. Third, as this study did not include detailed surgical outcome data, interpretation of postprocedural mRS should be limited to the procedural safety and early postoperative functional status, rather than long-term treatment efficacy. Herein, we identified tumor-specific preoperative intracranial tumor embolization complications and their risk factors, which were not analyzed in previous JR-NET studies. In the future, we intend to compare each tumor-specific study with previous JR-NET studies.
Conclusions
In JR-NET4, the incidence of the primary endpoint (mRS score 0-2 at 30 days after the procedure) was 86.9%, and the procedure-related complication rate was 4.8%. Target vessels other than ECA and general anesthesia were identified as risk factors for the development of complications. In meningiomas, a target vessel other than the ECA was a significant risk factor, whereas in hemangioblastomas, treatment with general anesthesia was a significant risk factor.
Ethical Approval
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration as well as its later amendments or comparable ethical standards. For this study type, formal consent was not required.
Disclaimer
Authors Koji Iihara, Shinichi Yoshimura, and Nobuyuki Sakai are members of the Editorial Board of Neurologia medico-chirurgica. These authors were not involved in the peer-review or decision-making process for this paper.
Conflicts of Interest Disclosure
All authors who are members and nonmembers of the Japan Neurosurgical Society (JNS) have registered self-reported COI disclosure statements through the website for JNS. Nobuyuki Sakai reports a research grant from Biomedical Solutions, Medtronic, Terumo, and TG Medical; lecturer's fees from Asahi-Intec, Biomedical Solutions, Daiichi-Sankyo, Kaneka, Medtronic, and Terumo; and membership on the advisory boards from Johnson & Johnson, Medtronic, and Terumo outside the submitted work. Other coauthors have no conflict of interest for this manuscript.
Acknowledgments
The JR-NET4 Study Group: Principal investigator; Nobuyuki Sakai, Kobe City Medical Center General Hospital, Kobe, Japan; Investigators; Koji Iihara and Hirotoshi Imamura, National Cerebral and Cardiovascular Center, Suita, Osaka, Japan, Akira Ishii, Juntendo University Graduate School of Medicine, Tokyo, Japan, Yuji Matsumaru, University of Tsukuba, Tsukuba, Ibaraki, Japan, Chiaki Sakai, Kyoto University, Kyoto, Japan, Tetsu Satow, Kindai University, Osaka-Sayama, Japan, Shinichi Yoshimura, Hyogo Medical University, Nishinomiya, Hyogo, Japan, and Certified Specialist of Japanese Society of Neuroendovascular Therapy.
Funding Statement
This study was supported in part by a research grant from the Japanese Society of Neuroendovascular Therapy and by the Kobayashi Foundation, Osaka, Japan.
References
- 1).Shah AH, Patel N, Raper DM, et al. The role of preoperative embolization for intracranial meningiomas. J Neurosurg. 2013;119(2):364-72. doi: 10.3171/2013.3.JNS121328 [DOI] [PubMed] [Google Scholar]
- 2).Przybylowski CJ, Zhao X, Baranoski JF, et al. Preoperative embolization versus no embolization for WHO grade I intracranial meningioma: a retrospective matched cohort study. J Neurosurg. 2021;134(3):693-700. doi: 10.3171/2020.1.JNS19788 [DOI] [PubMed] [Google Scholar]
- 3).Sakai N, Yoshimura S, Taki W, et al. Recent trends in neuroendovascular therapy in Japan: analysis of a nationwide survey―Japanese registry of neuroendovascular therapy (JR-NET) 1 and 2. Neurol Med Chir (Tokyo). 2014;54(1):1-8. doi: 10.2176/nmc.oa.2013-0197 [DOI] [PubMed] [Google Scholar]
- 4).Hishikawa T, Sugiu K, Hiramatsu M, et al. Nationwide survey of the nature and risk factors of complications in embolization of meningiomas and other intracranial tumors: Japanese registry of neuroendovascular therapy 2 (JR-NET2). Neuroradiology. 2014;56(2):139-44. doi: 10.1007/s00234-013-1300-4 [DOI] [PubMed] [Google Scholar]
- 5).Sugiu K, Hishikawa T, Murai S, et al. Treatment outcome of intracranial tumor embolization in Japan: Japanese registry of neuroendovascular therapy 3 (JR-NET3). Neurol Med Chir (Tokyo). 2019;59(2):41-7. doi: 10.2176/nmc.st.2018-0220 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6).Al-Mufti F, Gandhi CD, Couldwell WT, et al. Preoperative meningioma embolization reduces perioperative blood loss: A multi-center retrospective matched case-control study. Br J Neurosurg. 2023;37(1):67-70. doi: 10.1080/02688697.2021.1979191 [DOI] [PubMed] [Google Scholar]
- 7).Ampie L, Choy W, Lamano JB, et al. Safety and outcomes of preoperative embolization of intracranial hemangioblastomas: a systematic review. Clin Neurol Neurosurg. 2016;150:143-51. doi: 10.1016/j.clineuro.2016.09.008 [DOI] [PubMed] [Google Scholar]
- 8).Jumah F, AbuRmilah A, Raju B, et al. Does preoperative embolization improve outcomes of meningioma resection? A systematic review and meta-analysis. Neurosurg Rev. 2021;44(6):3151-63. doi: 10.1007/s10143-021-01519-z [DOI] [PubMed] [Google Scholar]
- 9).Schartz D, Furst T, Ellens N, et al. Preoperative embolization of meningiomas facilitates reduced surgical complications and improved clinical outcomes: a meta-analysis of matched cohort studies. Clin Neuroradiol. 2023;33(3):755-62. doi: 10.1007/s00062-023-01272-4 [DOI] [PubMed] [Google Scholar]
- 10).Iacobucci M, Danieli L, Visconti E, et al. Preoperative embolization of meningiomas with polyvinyl alcohol particles: the benefits are not outweighed by risks. Diagn Interv Imaging. 2017;98(4):307-14. doi: 10.1016/j.diii.2016.08.006 [DOI] [PubMed] [Google Scholar]
- 11).Yin Y, Li Y, Jiang Z, et al. Clinical outcomes and complications of preoperative embolization for intracranial giant meningioma tumorectomy: a retrospective, observational, matched cohort study. Front Oncol. 2022;12:852327. doi: 10.3389/fonc.2022.852327 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12).Akimoto T, Ohtake M, Miyake S, et al. Preoperative tumor embolization prolongs time to recurrence of meningiomas: a retrospective propensity-matched analysis. J Neurointerv Surg. 2023;15(8):814-20. doi: 10.1136/neurintsurg-2022-019080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13).Yoon N, Shah A, Couldwell WT, et al. Preoperative embolization of skull base meningiomas: current indications, techniques, and pearls for complication avoidance. Neurosurg Focus. 2018;44(4):E5. doi: 10.3171/2018.1.FOCUS17686 [DOI] [PubMed] [Google Scholar]
- 14).Rinaldo L, Brinjikji W. Dangerous extracranial-intracranial anastomoses: what the interventionalist must know. Semin Intervent Radiol. 2020;37(2):140-9. doi: 10.1055/s-0040-1709155 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15).Salaud C, Decante C, Ploteau S, et al. Implication of the inferolateral trunk of the cavernous internal CAROTID artery in cranial nerve blood supply: anatomical study and review of the literature. Ann Anat. 2019;226:23-8. doi: 10.1016/j.aanat.2019.07.004 [DOI] [PubMed] [Google Scholar]
- 16).Geibprasert S, Pongpech S, Armstrong D, et al. Dangerous extracranial-intracranial anastomoses and supply to the cranial nerves: vessels the neurointerventionalist needs to know. AJNR Am J Neuroradiol. 2009;30(8):1459-68. doi: 10.3174/ajnr.A1500 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17).Hishikawa T, Sugiu K, Murai S, et al. A comparison of the prevalence and risk factors of complications in intracranial tumor embolization between the Japanese registry of neuroendovascular therapy 2 (JR-NET2) and JR-NET3. Acta Neurochir (Wien). 2019;161(8):1675-82. doi: 10.1007/s00701-019-03970-w [DOI] [PubMed] [Google Scholar]
- 18).Ilyas A, Przybylowski C, Chen CJ, et al. Preoperative embolization of skull base meningiomas: a systematic review. J Clin Neurosci. 2019;59:259-64. doi: 10.1016/j.jocn.2018.06.022 [DOI] [PubMed] [Google Scholar]
- 19).Palavani LB, Andreão FF, de Abreu LV, et al. Assessing the efficacy and safety of hemangioblastoma embolization: a comprehensive systematic review and meta-analysis. J Clin Neurosci. 2023;117:104-13. doi: 10.1016/j.jocn.2023.09.021 [DOI] [PubMed] [Google Scholar]