Article Text
Abstract
Background and purpose The risk factors for adverse events for example, intraprocedural rupture (IPR), ischemic complication within 24 hours, and rebleeding after endovascular coil embolization for a ruptured cerebral aneurysm remain unclear. The aim of this study was to identify significant associated risk factors.
Methods We retrospectively evaluated data from 5358 patients listed on the Japanese Registry of Neuroendovascular Therapy 3 during 2010–2014.
Results IPR, ischemic complications, and rebleeding occurred during or after 221 (4.1%), 226 (4.2%), and 63 (1.2%) treatments, respectively. All of the adverse events were significantly associated with the patients’ poor outcomes at 30 days. Multivariate analysis revealed the factors independently related to these adverse events as follows: (1) for IPR: female sex, bifurcation type, <5 mm aneurysm, emergent surgery, local anesthesia, a balloon-assisted technique; (2) for an ischemic complication: poor World Federation World Federation of Neurological Societies (WFNS) grade, wide neck, and stent-assisted technique; (3) for rebleeding: poor WFNS grade, bifurcation type, wide neck, and body filling as the initial result.
Conclusions Knowledge of the risk factors of endovascular coil embolization and paying attention to them are essential for patients’ safe treatment and good outcomes.
- aneurysm
- coil
- subarachnoid
Statistics from Altmetric.com
Introduction
Endovascular coil embolization for cerebral aneurysms has become widespread throughout the world, including Japan. The effectiveness of coil embolization for ruptured cerebral aneurysms has already been established by two randomized trials: the International Subarachnoid Aneurysm Trial (ISAT)1 and the Barrow Ruptured Aneurysm Trial.2 Although the prognosis after endovascular treatment is better than that after surgical clipping, aneurysms displayed more rebleeding after embolization than following surgical intervention without achieving the main goal.3 Intraprocedural complications and rebleeding after coil embolization were reported to result in a poor outcome in patients with subarachnoid aneurysms according to the results of the Japanese Registry of Neuroendovascular Therapy (JR-NET) and JR-NET2.4 It is thought that preventing these events could improve the clinical results of coil embolization.
The JR-NET is a nationwide retrospective registration study, and the collected data from 2005 to 2006, 2007–2009, and 2010–2014 formed the JR-NET, JR-NET2, and JR-NET3, respectively.5 6 We then analyzed the data of coil embolization for ruptured cerebral aneurysms in the JR-NET3 study, and herein report the factors related to intraprocedural events and rebleeding that occurred within 1 month after embolization.
Methods
Altogether, 40 177 endovascular treatment cases during 2010–2014 were registered for the JR-NET3 study, among which 5935 (16.0%) underwent embolization for ruptured cerebral aneurysms. We then extracted 5358 treatments (90.3%) in which the detailed data necessary for our analysis were available.
Detailed data for each case included: (1) background of the patient, including age, sex, and modified Rankin Scale (mRS) before the onset of the ruptured cerebral aneurysm; (2) characteristics of the ruptured aneurysms based on (a) the World Federation of Neurosurgical Societies (WFNS) classification, (b) the site, size, and shape of the aneurysms, and (c) the use of preoperative antiplatelet drugs; (3) data concerning the treatment, including anesthesia, timing, use of bi-plane and three-dimensional (3D) rotational angiography, the strategy, the coil used, and the timing of heparin administration; and (4) results of the treatment (eg, using the Raymond Scale score) immediately after the procedure.
Each definition is explained in table 1. Locations of aneurysms were classified into seven sites: (1) anterior communicating artery (AcomA) and anterior cerebral artery (ACA); (2) posterior communicating artery (PcomA); (3) distal internal carotid artery (ICA), which included the ICA terminal and anterior choroidal artery; (4) proximal ICA (eg, paraclinoidal aneurysms); (5) middle cerebral artery (MCA); (6) basilar artery (BA); and (7) vertebral artery (VA) and posterior cerebral artery (PCA). Shapes of aneurysms were categorized as: (1) small size/small neck (maximum diameter <10 mm and neck size <4 mm or the ratio to maximum diameter and neck size ≥1.5); (2) small size/wide neck (maximum diameter <10 mm and neck size ≥4 mm or the ratio to maximum diameter and neck size <1.5); (3) large (maximum diameter 10–25 mm); (4) giant (maximum diameter ≥25 mm); and (5) non-saccular type. An ischemic event was defined as a new lesion on postoperative CT or MRI that induced an additional neurological symptom.
For the univariate and multivariate analysis, we divided age into >65 and ≤65; the WFNS classification into grades 1–3 and 4–5; aneurysm size into <5 mm and ≥5 mm; aneurysm shapes into small neck (<4 mm) and wide neck (≥4 mm); aneurysm locations into a bifurcation type (AcomA, MCA, BA) or a non-bifurcation type aneurysm; the day of treatment into within 24 hours and >24 hour; and initial result into complete occlusion or neck remnant, and body filling (BF).
We analyzed factors related to intraprocedural ischemic events, intraprocedural rupture (IPR), and rebleeding within 1 month after coil embolization for the overall aneurysms and the BF group, using JMP 10 software (SAS Institute, Inc, Cary, NC, USA). We used χ2 test and Fisher’s exact test for the univariate analysis and extracted factors with p<0.1. Then we performed a multivariate analysis, using logistic regression. A value of p<0.05 was considered to indicate statistical significance.
Results
Among 5337 endovascular coil embolizations excluding 21 cases in which we did not have detailed data, IPR and rebleeding were encountered in 221 (4.1%) and 63 (1.2%) patients, respectively. The 226 (4.2%) episodes of ischemia occurred within 24 hours after the treatment among 5329 embolizations, excluding 29 cases in which we did not have detailed data (table 1). Both IPR and ischemic events occurred simultaneously in 15 patients, and the number of patients who, after IPR and ischemia, suffered rebleeding were five and four, respectively. Thus, adverse events such as IPR, intraprocedural ischemia, and rebleeding were encountered in 486 cases (9.1%). All of these patients induced poor outcomes at the 30 day follow-up (table 2).
Univariate analysis of risk factors associated with adverse events
The univariate analysis determined that female sex, bifurcation type aneurysm size <5 mm, emergent surgery, and treatment within 24 hours were significantly related to the occurrence of IPR (table 3). Risk factors for ischemic episodes within 24 hours after the embolization were a poor-grade WFNS classification, aneurysm size <5 mm, wide neck aneurysm, treatment strategy, and BF as initial results, as well as the use of preoperative antiplatelet drugs (table 4). Rebleeding occurred at a more significant rate in patients with a poor WFNS classification, bifurcation type, a wide neck aneurysm, and BF as initial results (table 5).
Multivariate analysis of risk factors associated with adverse events
We extracted factors with p<0.1 by univariate analysis and performed a multivariate analysis that focused on IPR, ischemic complications, and rebleeding. The results are shown in tables 3–5. Female sex, bifurcation type, an aneurysm <5 mm, emergent surgery, local anesthesia, and balloon-assisted embolization were factors significantly associated with IPR. WFNS grades 1–3, wide neck aneurysm, and stent-assisted embolization were significantly related to the development of ischemia within 24 hours. The risk factors for rebleeding after embolization were WFNS grade 4–5, bifurcation type aneurysm, wide-neck aneurysm, and BF as the initial results.
Analysis of risk factors associated with rebleeding in the body filling group
Additionally, we evaluated the factors related to rebleeding in the BF group regarding the initial results. In the univariate analysis, bifurcation-type aneurysm was the only risk factor for rebleeding. We conducted the multivariate analysis and included local anesthesia, for which the significance was at p<0.1, as well as WFNS grade 4–5 and wide neck aneurysm, which were risk factors for all of the aneurysms. The results of the multivariate analysis were the same as seen with the univariate analysis with an added test, the odds ratio (OR), which was 0.004 (p=0.010) for the bifurcation-type aneurysm (table 6).
Discussion
Endovascular coil embolization for treating ruptured cerebral aneurysms has been widely used along with such devices as coils, balloon catheters,7 8 and intermediate catheters,9 as shown by the results of the ISAT trial1 and the Barrow Ruptured Aneurysm Trial.2 However, a few adverse events have occurred during and after coil embolization, with the incidence reported to be greater than that for unruptured cerebral aneurysms.10 In this study, IPRs, ischemic events, and rebleeding after embolization were observed in 4.1%, 4.2%, and 1.2% of cases, respectively, and these intraprocedural and post-procedural complications led to poor outcomes at the 30 day follow-up. As avoiding these adverse events improved the clinical results of endovascular treatment and expanded its indications, we evaluated the factors associated with IPR, intraprocedural ischemia within 24 hours, and rebleeding after embolization.
Although IPRs have resulted from various problems, such as coil protrusion and microcatheter perforation,11 the IPRs in this study appeared at a rate of 4.1%, which was similar to those using our previous JR-NET and JR-NET2 data, and higher than that of unruptured aneurysms based on our JR-NET and JR-NET2 data.12 While there was no significant correlation between an IPR and the prognosis in previous reports,13 14 IPR was a factor related to poor prognosis, which was similar to our previous JR-NET and JR-NET2 data. This difference may be because the timing of the data was slightly older, although balloon catheters are now able to access more distal areas than previously and reduce intraprocedural hemorrhage by the inflation around the aneurysm neck. In addition, the data showed that the location of the PCA is an independent predictor of IPR. Our analysis indicated that the factors associated with IPR were female sex, bifurcation type, <5 mm aneurysm, emergent surgery, local anesthesia, and use of a balloon-assisted technique. This is the first report in which female sex was found to be related to IPR. Komiyama et al reported that angiography within 6 hours of onset increased aneurysmal rebleeding.15 Thus, emergent surgery may be a risk factor for the same reason. A previous analysis16 showed that treating a ruptured aneurysm under local anesthesia was inappropriate, with unexpected movement of the patients and the micro-instruments likely the main reason. Another previous study17 showed that small aneurysms were a predictor of IPR, and our data agreed. This may be the greatest challenge for endovascular coil embolization. The reason balloon-assisted embolization was a related factor could be that we were inflating the balloon catheter at the time of the IPR.
Thromboembolic events such as IPR worsen the clinical outcomes of patients. A recent report18 stated that plasma D-dimer elevation associated with WFNS grades IV and V was a risk factor for thromboembolic events during endovascular embolization for ruptured aneurysms. This may be different from our results because ischemia was defined in this study as a symptom, not as an angiographic finding. Furthermore, it would be difficult to detect ischemia in a patient with severe subarachnoid hemorrhage. Coils may easily protrude during the embolization of a wide-necked aneurysm and are thought to increase the number of thromboembolic events.19 Although various adjunctive techniques (eg, use of a double catheter2 or a double-balloon technique,20 21) and new devices (eg, intracranial stents, an intrasaccular device) have recently been applied, it is a weak point that has not yet been resolved. A recent report22 stated that coiling with stent placement for ruptured wide-necked aneurysms did not increase perioperative procedure-related complications due to the improved skills of the interventionists and their strategies. The reason our result was different from their report may be that our data were registered from 2010 to 2014, making our results slightly more dated.
Although rebleeding is the most important factor in patient outcomes derived from analysis of the JR-NET and JR-NET2 data,4 related factors remain unclear. In the ISAT sub-analysis, the rebleeding rate within 30 days after the procedure was 1.9%, which was higher than that reported by a neurosurgical group.3 The rate in the current study was 1.2%, which was similar. Small aneurysms, an anterior communicating artery aneurysm, unilateral hypoplasia of the A1 segment, the presence of an intracerebral hematoma, high Fisher’s grading scale, aneurysmal outpouching, incomplete occlusion as the initial angiographic result, and adverse events during the procedure were reported to be significantly associated factors that could lead to rebleeding after coil embolization.23–26 We could not analyze unilateral hypoplasia of the A1 segment, the presence of an intracerebral hematoma, or aneurysmal outpouching due to data defects. The bifurcation-type aneurysm and incomplete occlusion as initial angiographic results were related factors, as previously reported, whereas small aneurysms, showing a high Fisher’s grading scale, and adverse events during the procedure were not risk factors. WFNS poor-grade and wide-necked aneurysms were newly recognized (in this study) significantly associated factors. Finally, in the BF group regarding the initial results, the bifurcation-type aneurysm was the only risk factor related to rebleeding within 30 days. We need to accumulate and analyze more detailed data in the future.
Study limitations
The most important drawback of the study is that it was a retrospective registry analysis. Additionally, the radiographic data were interpreted in a non-blinded manner, so “judgement calls”, such as the angiographic results, were obviously not precise measurements. Also, data such as the presence of intracerebral hematomas and aneurysmal blebs, which were evaluated in other reported studies, were not registered in this study. Thus, the evaluation of consistency with previous reports must be judged insufficient. The follow-up period was as short as 30 days to ensure there were fewer missing data. Thus, we could not evaluate the additional risk of delayed rebleeding after 30 days.
Conclusions
All adverse events, such as IPR, ischemic complications within 24 hours, and rebleeding after endovascular coil embolization for ruptured cerebral aneurysms, were significant factors associated with poor 30 day outcomes of patients. Factors related to IPR were female sex, bifurcation type, <5 mm aneurysm, emergent surgery, local anesthesia, and a balloon-assisted technique. Factors related to ischemic complications were a poor WFNS grade, a wide-neck aneurysm, and stent-assisted embolization. Factors related to rebleeding were a poor WFNS grade, bifurcation-type aneurysm, wide-neck aneurysm, and incomplete occlusion as an initial angiographic result. Preventing these adverse events should lead to improved endovascular treatment for ruptured cerebral aneurysms.
Acknowledgments
We thank Nancy Schatken, BS, MT(ASCP), from Edanz Group (www.edanzediting.com/ac), for editing a draft of this manuscript.
References
Footnotes
Collaborators JR-NET3 Study Group: Co-principal investigators: Nobuyuki Sakai, Kobe City Medical Center General Hospital, Kobe, Japan; Koji Iihara, Kyushu University, Fukuoka, Japan; Tetsu Satow, National Cerebral and Cardiovascular Center, Suita, Japan. Investigators: Masayuki Ezura, Sendai Medical Center, Sendai, Japan; Akio Hyodo, Dokkyo Medical University Saitama Medical Center, Koshigaya, Japan; Shigeru Miyachi, Aichi Medical University, Aichi, Japan; Susumu Miyamoto, Kyoto University, Kyoto, Japan; Yoji Nagai, Kobe University, Kobe, Japan; Kunihiro Nishimura, National Cerebral and Cardiovascular Center, Suita, Japan; Kazunori Toyoda, National Cerebral and Cardiovascular Center, Suita, Japan. Co-investigators: Toshiyuki Fujinaka, Osaka Medical Center, Osaka, Japan; Toshio Higashi, Fukuoka University, Fukuoka, Japan; Masaru Hirohata, Kurume University, Kurume, Japan; Akira Ishii, Kyoto University, Kyoto, Japan; Hirotoshi Imamura, Kobe City Medical Center General Hospital, Kobe, Japan; Yasushi Ito, Shinrakuen Hospital, Niigata, Japan; Naoya Kuwayama, Toyama University, Toyama, Japan; Hidenori Oishi, Juntendo University, Tokyo, Japan; Yuji Matsumaru, Tsukuba University, Tsukuba, Japan; Yasushi Matsumoto, Konan Hospital, Sendai, Japan; Ichiro Nakahara, Fujita Medical University, Aichi, Japan; Chiaki Sakai, Hyogo College of Medicine, Nishinomiya, Japan; Kenji Sugiu, Okayama University, Okayama, Japan; Tomoaki Terada, Showa University Fujigaoka Hospital, Kanagawa, Japan; Shinichi Yoshimura, Hyogo College of Medicine, Nishinomiya, Japan; and Certified Specialists of Japanese Society of Neuroendovascular Therapy.
Contributors HI contributed to analysis, interpretation of data, and preparation of the manuscript. All other authors reviewed it, and approved the final version, and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Funding This study was supported in part by a Grant-in-Aid (Junkanki-Kaihatsu H24-4-3) from the National Cerebral and Cardiovascular Center, Japan and by Hatazaki Foundation, Kobe, Japan.
Competing interests HI received Speakers’ Bureau/Honoraria from Medtronic Co. NS received Speakers’ Bureau/Honoraria from Otsuka Pharmaceutical Co, Stryker Co, Medtronic Co, Medico’s Hirata Co, and Biomedical Solutions Co, and research funding from Otsuka Pharmaceutical Co, Terumo Co, and Daiichi Sankyo Co. KI received Speakers’ Bureau/Honoraria from Otsuka Pharmaceutical Co, and research funding from Otsuka Pharmaceutical Co, Mitsubishi Tanabe Pharma Co, Kaneka Medix Co, Chugai Pharmaceutical Co, and Eizai Co.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available in a public, open access repository.