Article Text

Download PDFPDF

Original research
Anatomical results, rebleeding and factors that affect the degree of occlusion in ruptured cerebral aneurysms after endovascular therapy
  1. Eduardo Murias Quintana1,
  2. Alberto Gil Garcia1,
  3. Pedro Vega Valdés1,
  4. Hugo Cuellar2,
  5. Ángela Meilán Martínez1,
  6. Antonio Saiz Ayala1,
  7. Serafin Costilla Garcia1,
  8. Dolores Escudero Augusto3,
  9. Julio Cesar Gutierrez Morales4,
  10. Antonio López García4
  1. 1Department of Radiology—Interventional Neuroradiology, Hospital Universitario Central de Asturias, Oviedo, Asturias, Spain
  2. 2Department of Neurosurgery, LSU Health Sciences Center, Shreveport, Louisiana, USA
  3. 3Department of Critical Care, Hospital Universitario Central de Asturias, Oviedo, Asturias, Spain
  4. 4Department of Neurosurgery, Hospital Universitario Central de Asturias, Oviedo, Asturias, Spain
  1. Correspondence to Dr Eduardo Murias Quintana, Department of Radiology—Interventional Neuroradiology, Hospital Universitario Central de Asturias, Julián Clavería s/n, Oviedo, Asturias 33006, Spain; emuriass{at}


Object To present a series of ruptured cerebral aneurysms in consecutive non-selected patients treated with endovascular therapy, analyzing the initial degree of occlusion, its anatomical evolution mid-term and the variables that could statistically affect them.

Methods 251 aneurysms were first treated with coiling (embolization). 203 patients were followed up with conventional angiography for 6–8 months after the initial treatment and 182 were followed up with three-dimensional time of flight MR angiography at18–24 months. Postoperative and mid-term anatomical results were evaluated anonymously and independently using the modified Montreal Scale.

Results The initial rate of complete occlusion was 70.9%, with rates of neck remnants and aneurysm remants of 18.3% and 10.7%, respectively. The recurrence rate was 13% after 6 months and 2% between 6 months and 2 years. The rate of retreatment was 11%. Statistically, the variables that were found to be related to the initial degree of occlusion were the use of a remodeling balloon technique (p=0.012), the size of the aneurysm neck (p=0.044) and the size of the aneurysm (p=0.004). The recanalization rate at mid-term depended on the size of the aneurysm. Although aneurysms with partial occlusion initially tended to evolve to a worse degree of closure than those with complete occlusion initially, the relationship was not statistically significant (p=0.110).

Conclusions Embolized aneurysms can develop a worse degree of closure even when the initial occlusion is complete. The degree of occlusion depends directly on morphological factors and the use of balloon-assisted techniques. The recanalization rate at mid-term depends on the size of the aneurysm and probably on the density of the packing achieved with the initial treatment.

  • Aneurysm
  • MRI
  • Magnetic Resonance Angiography
  • Subarachnoid

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Since the publication of the International Subarachnoid Aneurysm Trial (ISAT) in 2002, most centers have accepted that ruptured brain aneurysms can be treated safely with embolization.1 ,2 The most important guidelines state that, if both surgical and endovascular techniques are feasible, the latter is the technique of choice.3 ,4

Several problems intrinsically associated with endovascular treatment have been described, including a higher rate of partially occluded aneurysms and a higher rate of recanalization compared with surgical treatment.3 Since the publication of the CARAT study and ISAT studies, we know that the recanalization rate and partial occlusion closure have a direct influence on the rebleeding rate.1 ,2 ,5–9

In studies without retreatment and with a high rate of partial occlusion closures such as ISAT, the rebleeding rate increased progressively from the first month up to 2 years. The retreatment of partial occlusion closures prevents rebleeding, as shown by the CLARITY study in which no rebleeding was observed after the first month and throughout the first year.1 ,8 Retreatment can be done by endovascular or surgery techniques.10 ,11

Another problem is the recanalization of aneurysms treated with endovascular techniques. Embolized aneurysms should therefore be followed long-term either by arteriography or, in recent times, by MR angiography (MRA). The aim is to detect a worsening in the degree of occlusion.12–15

It is therefore necessary to know the variables that influence the degree of initial occlusion of the aneurysm and to analyze the rate of worsening of the degree of occlusion and retreatment in order to try to minimize the rate of rebleeding and recanalization.

This study examined the anatomical and rebleeding results obtained in a series of consecutive unselected patients with ruptured cerebral aneurysms following initial embolization, with a minimum follow-up of 2 years. We compared our results with those in the literature and analyzed the variables related to the anatomical results in the short- and mid-term.

Materials and methods


An experienced neuroradiologist and an experienced neurosurgeon analyzed each aneurysm associated with a subarachnoid hemorrhage (SAH), and the endovascular option was chosen when both techniques were feasible or embolization was the better option. As reported in other related articles, the following aneurysms were not included: dissecting, fusiform, those associated with arteriovenous malformation and those previously treated with coils or clips. Variables such as age, past medical history (hypertension, diabetes and smoking) and aneurysm-related variables (location, neck size and dome:neck ratio) were recorded.

Treatment methodology

The majority of coils used were biological (Cerecyte; Micrus Endovascular Corporation, California, USA) and in a small percentage of cases a Guglielmi detachable coil (GDC) was used (Boston Scientific Neurovascular, Fremont, California, USA). When it was anatomically possible, treatment using the remodeling technique was chosen using Hyperglide or Hyperform balloons (ev3/Covidien). After the introduction of the first coil, 5000 units of heparin were administered and the patient was maintained coagulation-free for 24 h after the procedure using sodium heparin and prescribed antiplatelet treatment with 100 mg aspirin for 1 month.

Clinical assessment

All hospitalized patients were assessed using the Hunt and Hess scale, the Glasgow coma scale and the World Federation of Neurosurgical Societies (WFNS) scale. On discharge, a clinical assessment was carried out after 1 month and 1 year using the modified Rankin Scale (mRS).

The angiographic complications were separated into thromboembolisms due to the presence of a thrombus or a compromised space in a cerebral artery (independently of the percentage of flow reduction) and rupture of the aneurysm when contrast leakage was observed or coils were protruding into the subarachnoid space without contrast leakage.

The presence of vasospasm was assessed by clinical monitoring, transcranial Doppler and/or digital subtraction angiography (DSA) post-intervention. The patients were also monitored for stroke and hydrocephalus.

Anatomical assessment

The initial assessment of the degree of occlusion was investigated by DSA immediately after the treatment of the aneurysms with the consensus of two experienced neuroradiologists using the Montreal scale. This scale divides the degree of aneurysm occlusion into three groups: complete occlusion, neck remnants and aneurysm remnants.

The evolution of the degree of closure of the aneurysm was analyzed by angiography at 6 months and monitored with MRA 18–24 months after the initial treatment (time of flight 1.5 T, GE). The aneurysms were divided into three groups after initial DSA based on the consensus of two experienced neuroradiologists: improvement, no change or worsening. We identified those aneurysms that required retreatment and those which had rebled.

A statistical analysis was performed using the clinical (high blood pressure, age, gender and diabetes), anatomical (aneurysm and neck size, distance neck:dome and location) and treatment-related factors (degree of initial closure, use of assisted techniques and complications) to study the relationship with the degree of initial occlusion and the anatomical evolution of the aneurysms.

Statistical analysis

Continuous variables are represented using averages, SDs and minimum, maximum and average values. The continuous variables were compared using parametric tests (Student–Welch test or Welch–Robust test depending on the case) or non-parametric tests (Kruskal–Wallis) on variables that did not follow a normal distribution. The categorical variables are described using relative and absolute frequencies and the relationships between the categorical variables were measured using the exact χ2 test. Multivariate logistic regression analyses were carried out to establish which variables really had an influence on each prognosis. The qualities of the models were assessed using the area under the curve. p Values <0.05 were considered to be statistically significant.


Patient population

From 1 January 2005 to 31 December 2009, 721 patients were admitted to our center with SAH, of which 312 were diagnosed as having a cerebral aneurysm. After discounting non-saccular aneurysms, as previously mentioned, the sample was reduced to 273 aneurysms. After clinical assessment, 22 of these were chosen for surgical treatment: 14 because they had compressive hematoma with clinical instability and it was not possible to provide the necessary level of continuous patient care in our center (8 middle cerebral artery (MCA), 4 M2 MCA and 2 pericallosal), six due to coil instability (all dome neck <2 mm) and two because arterial cervical access was not possible. Thus, 251 aneurysms were included in the review (figure 1).

Figure 1

Anatomical evolution of the aneurysms following initial treatment (first row) in the DSA after 6–8 months (second row) and after MR angiography after 18–24 months (third row). DSA, digital subtraction arteriography; SAH, subarachnoid hemorrhage.

Aneurysm characteristics

Two hundred and twenty-six aneurysms were located in the anterior circulation (90%), with 28.2% in the internal carotid artery, 44.3% in the anterior communicating artery, 12.3% in the middle cerebral artery and 10% in the vertebrobasilar system. The average size of the aneurysms was 7.1 mm, with 32 aneurysms <3 mm (12.7%) and four giant aneurysms (1.6%). 92% of the aneurysms had a neck size <4 mm and the dome:neck ratio was >1.5 in 75.6% (table 1).

Table 1

Demographic and aneurysm characteristics

Intervention characteristics

The average period up to the intervention (time from bleeding to treatment) was 58.3 h, with a minimum of 2 h and a maximum of 19 days. Treatment was carried out in the first 72 h in 88% of the 221 patients. Assisted techniques were used in 146 patients (58.2%), including 143 patients (57%) with the remodeling technique. Stent-assisted embolization, double catheterization and parent vessel occlusion were each used in one case (0.4%).

Thromboembolic complications were observed in 22 patients (8.8%), leading to a neurological deficit in six cases (2.4%) and death in four patients (1.6%). Intraprocedural aneurysm rupture occurred in 11 patients (4.4%), which led to a worsening of the neurological condition in two patients (0.8%) and death in another two (0.8%). Rebleeding occurred in four patients (1.5%), which led to a worsening of the neurological condition in two patients (0.8%) and death in another two (0.8%). The permanent morbidity rate related to the procedure was 3.2% and the related death rate during the procedure was 2.4%.

Ninety patients (36%) developed hydrocephalus during follow-up imaging, and it was considered appropriate to use ventricular drainage in 82 patients (91.1%). Sixty-nine patients (27.5%) met the ultrasound criteria for cerebral vasospasms, which was clinically and/or angiographically compatible in all of them. Fifty-nine patients (23.5%) had cerebral ischemia in the follow-up neuroimaging. One hundred and seven patients complied with the post-intervention anticoagulant protocol (62.5%) and 150 patients complied with the post-intervention antiplatelet protocol (59.8%) (table 2).

Table 2

Patient distribution according to clinical grade on hospitalization on the WFNS scale over a series of different endovascular treatments

Initial aneurysm occlusion

Complete occlusion was achieved in 178 patients (70.9%), neck remnants were observed in 46 patients (18.3%) and aneurysm remnants in 27 patients (10.7%) (table 3). The variables that were found to be significantly related to the degree of occlusion were the use of a remodeling balloon (p=0.012), aneurysm neck size (p=0.044) and aneurysm size (p=0.004).

Table 3

Immediate anatomical results of our series compared with previous series

At 6-month follow-up

Four patients were retreated using endovascular techniques during the three first 3 months after treatment because of initial partial occlusion of the aneurysm, and subsequently they continued follow-up treatment. Two hundred and three patients underwent the first control arteriography; 35 died and 13 refused to do the control.

The degree of closure occlusion worsened in 36 patients (15 had complete closure, 5 neck remnants and 6 aneurysm remnants). Treatment remained stable in 146 patients (10 had neck remnants and 4 had aneurysm remnants after the initial treatment). An improvement in the degree of closure was observed in 31 patients, evolving to complete closure occlusion (20 had neck remnants and 11 had aneurysm remnants after the initial treatment).

At 2 year follow-up

Eighteen patients were retreated following the first angiographic control (12 were re-embolized and 6 were treated by surgery). One hundred and eighty-two patients continued monitoring with MRA after 2 years of the initial treatment (3 patients died). In four patients the degree of occlusion worsened (1 had stable neck remnants after 6 months and 3 had stable complete closure after 6 months). Treatment remained stable in the rest of the aneurysms.

Aneurysm retreatment and reopening rates

We therefore obtained a rate of worsening in the degree of occlusion of 13% (26/203) after 6 months and 2% (4/182) after 2 years and a retreatment rate of 11% (table 4).

Table 4

Comparison of our anatomical results after 2 years with those of the Fern meta-analysis and the CLARITY study

Variables that affect worsening in the degree of closure

A statistically significant relationship was found between aneurysm size and worsening of the degree of closure (p=0.010). The degree of closing tended to worsen more in aneurysms with a low degree of packing than in those with complete occlusion in the initial treatment (p=0.110).


To comply with the recommendations for large randomized trials in real situations and as observed in the CLARITY study, our results come from consecutive non-selected patients.12 Also, the distribution of our series is similar to other published series such as ISAT and CLARITY in terms of age, gender and patient history.

However, in our sample we had more MCA aneurysms, a higher percentage of microaneurysms, a much higher percentage of patients in a poor clinical condition and fewer posterior circulation aneurysms than in the ISAT trial. We can consider that our interventional complications and mRS score at discharge are within the standards.

Post-embolization anatomical results

For optimal aneurysm occlusion it has to be either closed completely or to have neck remnants left to preserve the origin of a branch.8 We obtained optimal occlusion in approximately 90% of our cases. These results are similar to those obtained in the CLARITY study. Although there are different anatomical classifications, we have compared the morphological results in coiled aneurysms with previously published series in table 3.8 ,16–21 Clinical factors such as the patient’s clinical condition at admission and postoperative complications were not statistically related to the initial degree of occlusion.

The degree of initial occlusion is directly influenced by aneurysm-related anatomical factors. Some aneurysms with a larger diameter or bigger neck (factors treated as continuous variables) tend to present initially with a lower degree of packing. This ratio is also maintained if we analyze aneurysms with a neck size >4 mm. The initial use of the remodeling balloon technique increases the degree of occlusion in the aneurysms, giving a statistically significant relationship between its use and complete closure. This assisted technique is therefore fundamental to achieving a suitable degree of occlusion initially. On the other hand, no significant relationship was found between the degree of initial closure and the location of the aneurysm, either in the anterior or posterior circulation.


In the mid-term we confirmed that some aneurysms tend to reopen (recanalyze) even when complete occlusion was achieved after the initial treatment. It is therefore necessary to monitor all embolized aneurysms after endovascular treatment, even if complete occlusion was achieved.

The recanalization rate of endovascular-treated aneurysms in previous series ranges from 10% to 20%; in our series the recanalization rate was 13% after 6 months and 2% between 6 months and 2 years. The tendency to recanalize therefore tapers off significantly with time, a finding that has already been described in some related articles.1 ,8 ,20 ,22–25

Moreover, partially treated aneurysms can develop complete occlusion, probably due to re-epithelialization of the neck. In our study the proportion was close to 15% and occurred mainly during the first 6 months after embolization.

We did not consider that the patient’s clinical condition at admission and the postoperative complications were related to worsening in the degree of occlusion. Other published studies have found a direct relationship with age, high blood pressure and smoking. In fact, these last two factors have become critical, being the cornerstone of any medical treatment when preventing recanalization of aneurysms. In our study we only recorded the presence of high blood pressure and smoking while the patient was in hospital care and not during follow-up, so only limited conclusions can be drawn in this regard.

We found a statistically significant relationship between aneurysm size and worsening in the degree of occlusion, with greater worsening over time in larger aneurysms than in small ones. This same conclusion has been reported in other studies.7 ,8

Some previous studies have found that aneurysms with a greater packing density after initial treatment are less likely to have worsening in the degree of occlusion than aneurysms with a low packing density. In our study there was no statistically significant relationship (p=0.110) due to the small sample size, with results similar to some described in other studies.8


One of the most significant findings was the absence of rebleeding after the first month and during the first 2 years, which is similar to the results obtained in the CLARITY study. In the ISAT study the rate of rebleeding was 1.2% in the first month and 7% between the first month and 2 years in the endovascular arm.1 ,6

We found that obtaining complete occlusion is key to avoiding rebleeding and to achieving results similiar to those obtained with surgery. The use remodeling balloon in acute phase in all cases or early retreatment with endovascular stent-assisted techniques will improve the results.

The rate of retreatment in our study was higher than in the ISAT study but the same as in the CLARITY study. This is probably explained by the larger number of aneurysms with optimal occlusion in our study and in the CLARITY study than in the ISAT study.

It is important to consider the role of surgery in the retreatment of partially treated aneurysms, especially in young patients, and endovascular stent-assisted techniques in order to reduce the rate of rebleeding and partially treated aneurysms. Prospective trials are needed to compare the two methods.


Although embolized aneurysms tend to have a higher rate of recanalization, even if complete occlusion is achieved after the initial treatment, the rate of early rebleeding is very low. Early rebleeding occurs in patients with aneurysm remnants, so early retreatment is essential in this group. Aneurysm recanalization occurs mainly in the first months after embolization, and long-term follow-up after 2 years is needed only in selected cases.

The degree of initial occlusion depends directly on morphological factors and the use of the balloon remodeling technique. The rate of mid-term reopening depends directly on the size of the aneurysm and probably on the packing density achieved in the initial treatment.

These data suggest that the evolution of endovascular therapy has steadily decreased the rate of rebleeding, increasing the complete occlusion of cerebral aneurysms. This has enabled us to achieve our objective of similar results between surgery and endovascular therapy in terms of rebleeding and recanalization.



  • Contributors EMQ: integrity of the study, study design, data collection, analysis and interpretation of the data, statistical treatment. literature research, text writing, critical revision and approval of the final version. AGG, PVV and HC: critical revision and approval of the final version. ÁMM: text writing, critical revision and approval of the final version. ASA: study design, text writing, critical revision and approval of the final version. SCM: study design, critical revision and approval of the final version. DEA: critical care, critical revision and approval of the final version. JCGM: study design, critical revision and approval of the final version. ALG: critical revision and approval of the final version.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics approval was obtained from the Ethics Committee of the Hospital Universitario Central de Asturias.

  • Provenance and peer review Not commissioned; externally peer reviewed.