Systematic Review of Transcatheter Arterial Embolization of AVM: Indications, Bleeding Complications, Cure Rate, and Long-Term Bleeding Risk
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.304101
Source of Support: None, Conflict of Interest: None
Keywords: Arteriovenous malformation, bleeding complications, cure rate, indications, transarterial embolization
Arteriovenous malformations (AVMs) are vascular malformations that result from birth defects involving the vessels of both arterial and venous origins, resulting in direct communications between the different size vessels or a meshwork of primitive reticular networks of dysplastic minute vessels, which have failed to mature to become “capillary” vessels termed “nidus”. These lesions are defined by shunting of high velocity, low resistance flow from the arterial vasculature into the venous system in a variety of fistulous conditions. Prevention of hemorrhage is still the main goal of brain AVM treatment. Transarterial AVM embolization is used to eliminate a possible cause of hemorrhage before surgery or radiosurgery treatment or to obliterate small malformations.,,,, N-butyl cyanoacrylate (NBCA) and Onyx have been extensively used for transarterial AVM embolization.,,,,,,,,,,, Variations in endovascular practice exist, due to referral practices, availability of technical equipment and expertise, personal preference, and absence of random prospective tests comparing treatment modalities. Outcomes following transarterial AVM embolization vary greatly between studies.,,,,,,,,,,,, To our experience, AVM embolization seldom resulted in ischemic injury that produces death or disabling morbidity, bleeding complications during or post AVM embolization are the critical problem. Conflicting reports in the literature provide evidence for a decreased risk of hemorrhage, an unchanged risk of hemorrhage, and an increased risk of hemorrhage. What were the obliteration and post-embolization bleeding rates in these patients? It would be interesting to assess the effectiveness of transarterial AVM embolization and how it impacts the natural history. In this review, we assess the proportion of patients that have bleeding complications, cure rates, and long-term bleeding events for each of the indications (curative monotherapy, pre-SRS, and pre-surgery) for AVM embolization.
A systematic review search was done for all available published studies on PubMed by May 2016. Mesh terms included in the search were “intracranial arteriovenous malformations”[Mesh] AND “Embolization, Therapeutic”[Mesh]. No publication date or publication status restrictions were imposed. Studies were selected based on the following criteria: (1) clinical studies (including Spetzler Martin I-V); (2) I-butyl cyanoacrylate (IBCA)/NBCA or Onyx exclusively or predominantly used; and (3) outcomes: hemorrhagic events related to the embolization procedures. A secondary outcome was the complete obliteration rate of AVMs with embolization alone and long-term bleeding risk. Exclusion criteria included: (1) studies reporting on silk or particles; (2) series <40 cases; (3) studies not reporting on outcomes; and (4) duplicated studies from a single center conducted in the same periods. To enhance the comprehensiveness of our review, references in the reviewed studies were perused and incorporated if our inclusion criteria were met.
The data elements extracted from each study included: (1) number of patients; (2) mean age; (3) gender; (4) study period; (5) NBCA/IBCA/Onyx used; (6) mean embolization sessions; (7) percentage of Spetzler–Martin grades IV and V; (8) percentage of hemorrhagic presentation; (9) treatment indications and (10) study areas. The patients were screened for bleeding. The measured results included bleeding complications related to embolization, cure rate, and long-term bleeding risk. Long-term bleeding risk was intracranial hemorrhage beyond 30 days of embolization. Bleeding complications were defined as bleeding events that occurred within 30 days after treatment. The proportion of successfully obliterated was defined based on the report of literature.
Statistical heterogeneity of the studies was measured using the I2test; levels above 50% were considered significant heterogeneity. Funnel plots were used to assess if there was publication bias. Because heterogeneity was found to be high among studies, event rates were pooled across studies using random-effects meta-analysis with 95% CI. Poisson regression analysis was used to investigate the associations of the middle of the study duration on the cure rate and bleeding complications.
The factors of the number of patients (<100 patients, >100 patients), NBCA/Onyx, embolization sessions (<2, ≥2), percentage of Spetzler–Martin grades IV and V (<20%, 20%–40%, and >40%), treatment modality (monotherapy, pre-surgical, and pre-SRS embolizations) and study areas (Europe, Asia, and North America) were analyzed by χ2 test. The subgroup categories were defined by document cluster analysis to decrease the heterogeneity. The analysis was conducted using the “meta” and “metafor” packages of R software (version 3.4.0, R Foundation for Statistical Computing Vienna, Austria). All statistical tests were two-sided and statistical significance was defined as P < 0.05.
A total of 1380 studies and abstracts were found in the primary search. After exclusion of duplicated, nonhuman, and irrelevant studies by abstract reviewing, 71 studies were selected for full review; 27 studies met the selection criteria. Three additional studies were found by considering the reference lists of the previously mentioned studies [Figure 1]. A total of 30 studies were included in the analysis [Table 1]. Two studies were prospective, and the remainder was retrospective studies with a total of 5013 patients (2418 patients of monotherapy, 2033 patients of pre-surgical embolization, and 562 patients of pre-SRS embolization).
The random effect model was performed to evaluate the cure rate of transarterial AVM embolization. The heterogeneity test was significantly positive (I2 = 94%) [Figure 2] and [Figure 3]. The use of egger's test to detect the bias was significant (P = 0.024) [Figure 4]. The overall cure rate was 18%. Curative embolization was achieved in 24% of monotherapy, 10% of pre-surgical, and 16% after pre-SRS [Figure 2]. After cluster analysis, the heterogeneity in the monotherapy group mainly comes from two studies,, but the heterogeneity in pre-surgical and pre-SRS groups was still significantly positive.
Onyx is associated with a higher cure rate than NBCA and IBCA (χ2 = 18.56, P < 0.005). The cure rate was 25% in small studies (<100 patients) and 24% in large studies (>100 patients) (χ2 = 5.73, P < 0.025). The random effect model was performed after deleting the 2 heterogeneous studies mentioned above., In the Onyx group, the heterogeneity test was I2 = 0%; the cure rate was higher of 25% (χ2 = 6.25, P < 0.025) [Figure 5]. The cure rate was higher in Europe 26% (χ2 = 11.17, P < 0.005). The high proportion of Spetzler-Martin grade IV + V was associated with a lower cure rate (χ2 = 6.25, P < 0.025) [Figure 6]. Multiple session (≥2) was associated with a higher cure rate of 23% (χ2 = 10.53, P < 0.005). Two sessions seem the most appropriate strategy slightly increasing the cure rate and not increasing the bleeding complication rate in random-effect analysis. In Poisson regression analysis, it is found that the advances of techniques did increase the cure rate of transarterial AVM embolization along with the middle year of the study duration (P = 0.03) [Figure 7].
The overall cure rate of pre-surgical embolization was 10% and the heterogeneity is significant (I2 = 82%). The cure rate was 17% in small studies (<100 patients) and 8% in large studies (>100 patients) (χ2 = 12.75, P < 0.005). Onyx has a higher cure rate of 17% than 11% of NBCA (χ2 = 12.75, P < 0.005) [Figure 8].
The overall cure rate of pre-SRS embolization was 16% and the heterogeneity is significant (I2 = 93%).
The overall bleeding complication rate of transarterial embolization was 5% (I2 = 77%). Case fatality of bleeding complication was 31% (I2 = 47%). Pre-SRS embolization was associated with an increased bleeding complication rate of 8% (χ2 = 13.83, P < 0.005) than 6% of monotherapy and 4% of pre-surgical embolization. Onyx has a higher bleeding complication rate of 8% (χ2 = 8.27, P < 0.005). The bleeding complication rate was lower in Asia (4%; χ2 = 30.65, P < 0.005) than that of Europe (7%) and North America (5%). The small studies reported a higher bleeding complication rate of 7% than large studies of 4% (χ2 = 8.99, P < 0.005). Multiple session (≥2) 5% (χ2 = 1.64, P > 0.1) and high proportion of Spetzler-Martin grade IV + V 6% (χ2 = 0.44, P > 0.5) were not associated with a higher bleeding complication rate. The advances of techniques did not decrease the bleeding complication rate of transarterial AVM embolization (P = 0.19) [Figure 9].
Long-term bleeding risk
The long-term bleeding risk after transarterial AVM embolization was reported in 8 studies and ranged from 1.1% to 2.7%.,,,,,,, The mortality during follow-up was reported in 7 studies and ranged from 0.6% to 8%.,,,,,, Case fatality associated with long-term bleeding was not achieved because it was not known how many patients were followed, and it was not possible to allocate weights for each study.
The most appropriate embolic agents for primary AVM embolization include Onyx and NBCA.,,,,,,,,,,, Due to technical advancements, there is a trend towards an increase in obliteration rates in both NBCA and Onyx with time [Figure 10]. However, the bleeding complication rate over time is not decreased probably caused by the aggressive embolization and increasing the number of patients treated with embolization. Possible bleeding complications have been described more meticulously in more recent studies.,,,, All 3 treatment modalities had considerable risks of bleeding complications. In this systematic review of brain AVM treatment, we found that Onyx was associated with a higher cure rate and bleeding complication rate.
Although the annual rate of bleeding from untreated AVMs has been reported to be 2% to 4%,,, it is not suitable for all AVMs since some AVMs are more prone to rupture. The bleeding complication was 5.0%. Fatal bleeding was observed during the follow-up period in addition to therapeutic bleeding complications. These unfavorably high bleeding risks in treated AVMs in comparison with the natural course of untreated AVMs should raise serious concern. Pre-SRS AVM embolization was associated with a higher bleeding complication rate. This higher bleeding complication rate may in part be explained by the fact that aggressive size-reducing and fistula compartment embolization. Based on observational studies, there are no AVM characteristics that are hypothesized to predispose to bleeding complications in embolized AVMs.,,,,, This study has enhanced our understanding of AVMs and the evolution of their treatment. The clinical, morphological, and physiological characteristics of AVMs varied greatly between monotherapy, pre-surgery, and pre-SRS cases. Ruptured and superficial AVMs predispose to embolization as monotherapy or pre-surgery and unruptured, small, and deep AVMs predispose to pre-SRS embolization.
It is apparent that intention to AVM embolization outcomes is very difficult to analyze from retrospective studies in which the approach for some of the cases may not have been defined until after the embolization procedure. For example, some of the patients in the SRS and surgical group may have had initial intent of cure by embolization monotherapy. Therefore, the retrospective defining of these treatment groups may not reflect a prospective intention to treat these treatment groups. The evident bias, that failure of curative embolization will be followed by another treatment cannot be easily accounted for. Furthermore, a complication at the early stage may have precluded progress to further treatment. This may explain the lack of evident additional risk from multiple embolization procedures. There are situations where a specific goal could be to reduce perceived increased bleeding risk in the presence of clearly identifiable cause of the bleed such as an associated aneurysm or pseudo-aneurysm. The data did not provide information in this respect.
An important limitation of our meta-analysis is that the large majority of studies were retrospective and without independent outcome assessment. It is important to point out that characteristics of the brain AVM direct the choice for treatment modality and none of the data reported herein are from randomized controlled trials. The selection of patients for certain treatment options is likely to have been influenced by patient and brain AVM characteristics and, therefore, hampers comparison of treatments. The pathological process of AVM is also heterogeneous. The heterogeneity test was significantly positive assuming the observed estimates of treatment effectiveness can vary across studies. In addition, we may have excluded studies that may be important but did not fulfill all inclusion criteria. Publication bias is a widespread problem that may seriously distort attempts to estimate the effect under investigation. These findings suggest that conclusions based only on a review of published data should be interpreted cautiously, especially for observational studies. Improved strategies are needed to identify the results of unpublished as well as published studies. The strength of our meta-analysis is that we have an over review of transarterial AVM embolization.
Onyx has increased the cure rate and bleeding complication rate of AVM embolization. Multiple sessions are associated with a higher cure rate and do not increase the bleeding complication. Pre-surgical AVM embolization is a safe strategy. Current aggressive size-reducing and fistula-eliminating pre-SRS emblization has a higher bleeding complication rate. There are publication biases between small and large studies.
This work was supported by Beijing Municipal Administration of Hospitals Incubating Program (PX2020039), Beijing, China.
Informed consent was obtained from all individual participants included in the study.
Financial support and sponsorship
This work was supported by Beijing Municipal Administration of Hospitals Incubating Program (PX2020039), Beijing, China.
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9], [Figure 10]