Endovascular Treatment of Intracranial Aneurysms With Flow Diverters: A Meta-Analysis

Flow-diverter devices are new, important tools in the treatment of intracranial aneurysms.¹ Several single- and multicenter studies have demonstrated acceptable rates of aneurysm occlusion, morbidity, and mortality for patients treated with flow diverters.^2–30 These devices are being deployed in greater numbers of patients with more complex aneurysm morphologies and locations.^{9,12,18,22,25,29} With increasing experience, some of the limitations and unexpected complications of flow diverters have been recognized. These include intraparenchymal hemorrhage (IPH), postprocedural subarachnoid hemorrhage (SAH), as well as ischemic stroke.⁸

Improved understanding of safety and efficacy profiles associated with flow-diverter treatment of intracranial aneurysms would help guide practitioners in selection and follow-up of patients treated with these devices. We conducted a systematic review and meta-analysis of the literature regarding aneurysmal occlusion rates and procedure-related complication rates for intracranial aneurysms treated with flow diverters.

A comprehensive review of the literature was performed using the keywords “Intracranial aneurysm”, “divert”, “diversion”, “silk”, “pipeline,” and “pipeline embolization device” to search Pubmed, Ovid Medline, Ovid EMBASE, Scopus, and Web of Science database. Inclusion criteria were the following: English language, >5 patients, studies published between January 2005 and September 2012, and data on postoperative complications and aneurysmal occlusion rates. The exclusion criteria were the following: case reports, in vitro or cadaveric studies, review articles, guidelines, technical notes, and disaster series (series in which all patients were selected because of certain major complication).

The electronic search was supplemented by contacting experts in the field and reviewing the bibliographies of included studies for relevant publications. Abstracts, methods, results, figures, and tables of full text were searched for data on aneurysmal occlusion rates, procedure-related morbidity and mortality, and procedure-related complications.

Aneurysm occlusion was defined as complete occlusion at 6 months. We studied the effect of aneurysm size and aneurysmal occlusion rates, stratifying aneurysms as small (<10 mm), large (10 mm ≥aneurysm size ≤25 mm), or giant (>25 mm).

Procedure-related complications were stratified as total, early (within ≤30 days), and late (>30 days). Complications studied included total, early, and late IPH; total, early, and late ischemic stroke; total perforator infarction; and total, early, and late SAH. We examined the association between aneurysm size (small versus large/giant) and aneurysm location (anterior versus posterior) and the total rates of each of the studied complications.

Our meta-analysis demonstrated high occlusion rates for aneurysms treated with flow diverters, irrespective of aneurysm size. However, we also demonstrated that the complications associated with flow diverter treatment are not negligible, with morbidity and mortality rates of 5% and 4%, respectively. The safety of flow diversion in small aneurysms was superior to that of large aneurysms, with the latter associated with higher rates of both ischemic infarction and SAH. Higher morbidity in larger aneurysms may relate to the technical challenges as well as the inherent instability of these lesions. We did not find any specific aneurysm type associated with higher rates of IPH but did find an alarmingly high association between perforator infarction and posterior location of intracranial aneurysms. These findings suggest that practitioners must be judicious in selecting candidates for flow-diverter therapy, especially for large or posterior circulation aneurysms.

Published aneurysmal complete occlusion rates are often variable, ranging anywhere from 55% to 95%.^4,24 In combining aneurysmal occlusion rates from 29 studies, our meta-analysis provides more representative data on aneurysmal occlusion rates than any single study. Furthermore, in studying such a large sample, we had more power to detect differences in aneurysmal occlusion rates by size. This lends greater validity to our finding that aneurysmal occlusion rates were high regardless of size. This finding is extremely important because the current dogma in endovascular intracranial aneurysm treatment is that smaller aneurysms have better occlusion rates than larger aneurysms.³⁴ The finding that even giant aneurysms have such high occlusion rates provides important implications for those looking to stem the rates of aneurysm recurrence among this population.

Among larger studies, mortality rates have ranged from 0% to 7%,^2,28 whereas morbidity has ranged from 0% to 12%.^2,13 Our meta-analysis provides more representative data on morbidity, mortality, and complication rates associated with flow-diverter treatment. Another advantage of this meta-analysis is its increased power to detect differences in complication rates by aneurysm size and location, thus allowing practitioners to risk stratify potential patients.

One of the unanticipated and most feared complications of flow diverters is aneurysm rupture after treatment with these devices. The overall incidence of this complication is unknown, although it is thought to be low and critically related to the complexity of the treated aneurysm. Our meta-analysis suggests that SAH from delayed aneurysm rupture occurs in ≈4% of patients treated with flow diverters, with significantly higher rates among patients with large and giant aneurysms. Aneurysm rupture at ≥1 month postoperative was a relatively rare occurrence (2% of cases). Nonetheless, concerns regarding delayed rupture are so serious that Balt Extrusion issued a medical device alert instructing practitioners not to use the Silk flow diverter without coils owing to the potential for patient death.¹⁵ Our study emphasizes that postoperative SAH is a real and significant complication of flow-diverter patients, especially for those with large or giant aneurysms. It is not known at this point whether the practice of using endovascular coils in association with flow diverters in the treatment of larger aneurysms has resulted in a decreased incidence of this devastating complication.

IPH not associated with aneurysm rupture is another dreaded and poorly understood complication of flow-diverter treatment. Previous studies have reported rates ranging from 0% to 10% for this complication.^2,4,8,19 We demonstrated a 3% rate of IPH among flow-diverter patients. The mechanism for IPH is unknown. Hemorrhagic transformation of ischemic stroke, hemodynamic alteration from flow-diverter placement, and dual antiplatelet therapy are proposed mechanisms.⁸ In our study, neither aneurysm size nor location was associated with IPH rate. All studies included both pre operative and postoperative dual antiplatelet therapy; thus, we could not examine any independent role this might play in IPH formation. Further studies are needed to determine the ultimate cause of IPH.

Ischemic stroke and perforator infarction are well-described complications of flow-diverter treatment.^35,36 We demonstrated an ischemic stroke rate of 6%, with higher rates in posterior circulation aneurysms and large/giant aneurysms. Ischemic stroke is thought to result from thrombus formation along the stent wall, leading to stent occlusion, parent artery occlusion, or distal thromboembolic events.^35,36 The higher rates of ischemic stroke among patients with large/giant aneurysms may be related to the fact that, these aneurysms likely required more flow-diverter devices to achieve successful occlusion and may have been subject to longer operation times. Intraoperatively, acute thrombus formation can be mitigated by prompt injection of Abciximab.²⁷ However, in the long term, it is difficult to reduce the risk of thromboembolic events associated with flow-diverter treatment.

Perforator infarction in our meta-analysis was not uncommon and often led to devastating consequences. Perforator vessels in the posterior circulation are at particularly high risk for infarction relative to those in the anterior circulation. This is likely because of the delicate perfusion and lack of collaterals to brain stem structures. Many case series of posterior circulation aneurysms treated with flow diverters have demonstrated this fact.^12,21,25 Thus, given the relatively high rate of this complication and the devastating consequences of brain stem infarction, treatment of posterior circulation aneurysms in which perforator vessels could be involved should be performed only when absolutely necessary. When possible, deconstructive technique (parent vessel occlusion with flow reversal) should be the first consideration for treatment of large or giant basilar artery aneurysms.

This study has various limitations. Ecological bias (ie, comparisons are made across studies and not within studies), presence of publication bias, and statistical heterogeneity are limitations that affect all meta-analyses. Our study also has limitations because of the methodologic limitations of included studies. A majority of the included studies were retrospective case series. No prospective studies included were randomized or included control groups. Many of the included studies had a small sample size and incomplete follow-up data. Because a majority of previously published studies did not stratify outcomes based on important variables, such as patient demographics, duration of antiplatelet therapy, aneurysm rupture status, aneurysm subtype (secular versus fusiform), number of devices deployed, use of concomitant coiling, and previous aneurysm treatment status, we were unable to control for these findings in our analysis. Efficacy of flow-diverter treatment in treating compressive symptoms was rarely reported in the included studies; thus, we were unable to determine the efficacy of flow diverters in treatment of these symptoms. Our findings represent a very wide spectrum of aneurysms and clinical presentations, and thus our findings cannot be applied to determine the risks and complications associated with treating individual patients or different subgroups of aneurysms with flow diverters. Given the heterogeneity of the patients and aneurysms in our study, our data cannot be used to compare the efficacy of flow-diverter technology versus treatments, such as coiling and clipping. Therefore, the overall quality of evidence (strength of inference) presented in this systematic review is considered to be low.