Article Text

Download PDFPDF

Original research
Predicting intraprocedural rupture and thrombus formation during coiling of ruptured anterior communicating artery aneurysms
  1. Lianghao Fan1,
  2. Boli Lin2,
  3. Ting Xu2,
  4. Nengzhi Xia2,
  5. Xiaotong Shao2,
  6. Xianxi Tan3,
  7. Ming Zhong3,
  8. Yunjun Yang2,
  9. Bing Zhao3,4
  1. 1Department of Interventional Radiology, The First Affiliated Hospital, Wenzhou Medical University, Wenzhou, China
  2. 2Department of Radiology, The First Affiliated Hospital, Wenzhou Medical University, Wenzhou, China
  3. 3Department of Neurosurgery, The First Affiliated Hospital, Wenzhou Medical University, Wenzhou, China
  4. 4Department of Neurosurgery, Mayo Clinic, Rochester, Minnesota, USA
  1. Correspondence to Dr Bing Zhao, Department of Neurosurgery, or Dr Yunjun Yang, Department of Radiology, The First Affiliated Hospital, Wenzhou Medical University, Nanbai Xiang Town, 325000 Wenzhou, China; drzhaobing{at}yahoo.com, wzfskyyj2011{at}163.com

Abstract

Background Intraprocedural rupture and thrombus formation are serious complications during coiling of ruptured intracranial aneurysms, and they more often occur in patients with anterior communicating artery (ACoA) aneurysms.

Objective To identify independent predictors of intraprocedural rupture and thrombus formation during coiling of ruptured ACoA aneurysms.

Methods Between January 2008 and February 2015, 254 consecutive patients with 255 ACoA aneurysms were treated with coiling. We retrospectively reviewed intraoperative angiograms and medical records to identify intraprocedural rupture and thrombus formation, and re-measured aneurysm morphologies using CT angiography images. Multivariate logistic regression models were used to determine independent predictors of intraprocedural rupture and thrombus formation.

Results Of the 231 patients included, intraprocedural rupture occurred in 10 (4.3%) patients, and thrombus formation occurred in 15 (6.5%) patients. Patients with smaller aneurysms more often experienced intraprocedural rupture than those with larger aneurysms (3.5±1.3 mm vs 5.7±2.3 mm). Multivariate analysis showed that smaller ruptured aneurysms (p=0.003) were independently associated with intraprocedural rupture. The threshold of aneurysm size separating rupture and non-rupture groups was 3.5 mm. Multivariate analysis showed that a history of hypertension (p=0.033), aneurysm neck size (p=0.004), and parent vessel angle (p=0.023) were independent predictors of thrombus formation. The threshold of parent vessel angle separating thrombus and non-thrombus groups was 60.0°.

Conclusions Ruptured aneurysms <3.5 mm were associated with an increased risk of intraprocedural rupture, and parent vessel angle <60.0°, wider-neck aneurysms, and a history of hypertension were associated with increased risk of thrombus formation during coiling of ruptured ACoA aneurysms.

  • Aneurysm
  • Coil
  • Complication
  • CT Angiography

Statistics from Altmetric.com

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.

Introduction

Intraprocedural rupture and thrombus formation are serious complications of coiling of ruptured intracranial aneurysms and increase morbidity and mortality.1–3 These complications more often occur in patients with ruptured aneurysms and anterior communicating artery (ACoA) aneurysms.4–9 However, no study has focused on predictors of intraprocedural rupture and thrombus formation during coiling of ruptured ACoA aneurysms. Several studies have proposed that aneurysm morphology is an important risk factor for intraprocedural rupture or thromboembolic complication.6 ,8 ,10 CT angiography (CTA) has been proposed as a primary examination tool for the investigation of patients with ruptured aneurysms, and most patients with ruptured aneurysms undergo CTA before treatment decision-making.11 ,12 Therefore, a thorough knowledge of aneurysm morphologies based on CTA associated with these complications may help to reduce the risk of coiling.

We conducted a retrospective analysis of consecutive patients with ruptured ACoA aneurysms treated with endovascular coiling. Aneurysm morphological parameters were re-measured using reconstruction of CTA images. The purpose of this study was to identify independent predictors of intraprocedural rupture and thrombus formation during coiling of ruptured ACoA aneurysms.

Methods

Study design

This study was approved by the institutional review board. From January 2008 to February 2015, 254 consecutive patients with 255 ACoA aneurysms were treated with coiling in our hospital. Ruptured aneurysms amenable to both clipping and coiling were considered for coiling except for a large intracerebral hematoma requiring surgery and hematoma evacuation. The ACoA aneurysms with complex morphologies and very small size were considered for surgical clipping. Stent-assisted coiling was considered in aneurysms with an unfavorable morphology (neck size ≥4 mm or dome/neck ≤2.0). Using the electronic medical record system, we collected the following data: age, sex, medical history, clinical presentation, radiologic examination data on CT and CTA, procedure report, and clinical outcomes assessed with the Glasgow Outcome Score at discharge. In this study, 19 patients with poor quality CTA images which could not be reconstructed to measure aneurysm morphology, three patients with unruptured aneurysms, and one patient for whom coiling had failed because of procedure-related dissection of A1, were excluded. Finally, 231 patients with ruptured aneurysms were successfully treated with coiling and were included.

Definitions of intraprocedural rupture and thrombus formation

All patients were treated under general anesthesia and systemic heparinization after femoral sheath placement. Activated clotting time was maintained at two to three times the baseline value. Intraprocedural rupture was defined as angiographic contrast extravasation. A loading dose of 450 mg clopidogrel was administered either by a nasogastric tube or rectally before 2 h of emergency stent deployment. Thrombus formation was defined as angiographic filling defects in the region of the aneurysms or the distal anterior cerebral artery beginning at the ACoA. Intraprocedural angiograms were reviewed by both an independent interventional neuroradiologist and neurosurgeon, who were not involved in the treatment. Detailed medical records were also reviewed. Patients were grouped into intraprocedural rupture and non-rupture groups, and thrombus formation and non-thrombus groups.

Treatment protocol for intraprocedural complications

If intraprocedural rupture occurred, the effect of heparin was immediately reversed with protamine, further aneurysm coil deposition was attempted if safely possible, and blood pressure was rapidly controlled. If thrombus formation occurred, the heparin effect was not reversed after the procedure and selective injection of urokinase was administered after the coiling before 2010. A dose of 10 000–15 000 IU/mL/min was continuously administered and angiography was repeated every 10–15 min. After 2010, intravenous infusion of the glycoprotein IIb/IIIa antagonist, tirofiban, was carried out after coiling. A loading dose of 0.4 µg/kg/min was administered for about 30 min, and a dose of 0.1 µg/kg/min was maintained. Subcutaneous injection of the low molecular weight heparin, nadroparin, at a dose of 0.4 mL, was carried out twice daily for 3 days. Aspirin (100 mg) and 75 mg of clopidogrel were administered daily for 12 weeks if patients were treated with stent-assisted coiling.

CTA image acquisition and reconstruction

A workstation (V.4.6; GE Medical Systems) was used to reconstruct the three-dimensional images regenerated from the thin-slice data retrieved from the archives. All morphologies were measured by two neuroradiologists familiar with CTA reconstruction and measurement, and their mean values were used for the analysis. CTA image data were acquired from a 16-channel multidetector CT scanner (General Electric Medical Systems, Milwaukee, Wisconsin, USA) with a section thickness of 1.25 mm and a reconstruction interval of 0.625 mm, a 64-channel multidetector CT scanner (General Electric Medical Systems) with a section thickness of 0.625 mm and a reconstruction interval of 0.625 mm, and a 320-detector row CT scanner (Toshiba Medical Systems, Japan) with a section thickness of 0.5 mm and a reconstruction interval of 0.5 mm.

Measurement of aneurysm morphologies

Aneurysm size, aneurysm height, perpendicular height, aneurysm neck size, vessel size, aspect ratio, size ratio, flow angle, aneurysm angle, parent vessel angles, A1 segment configuration, and aneurysm projections have been described previously.13 ,14 Aneurysm size was the largest cross-sectional diameter of the aneurysm. Aneurysm height was measured between the center of the aneurysm neck and the longest distance to the aneurysm dome. Aneurysm perpendicular height was the largest perpendicular distance from the aneurysm neck to the aneurysm dome. Aspect ratio was the ratio of the maximum perpendicular height to the aneurysm neck size. Flow angle was the angle between the maximum aneurysm height and the parent artery. Aneurysm angle was the angle between the aneurysm neck and the maximum aneurysm height, and parent vessel angle was the angle between the parent artery and the aneurysm neck. The parent artery was defined as the distal part of the A1 segment. The mean values of each angle were measured when the aneurysms occurred at a vessel junction (see online supplementary figure e1).

Statistical analysis

Statistical analyses were carried out with IBM SPSS V.22.0 (IBM SPSS, Armonk, New York, USA). Continuous variables were presented as means±SD, and categorical variables were presented as frequency (percentage). Univariate regression analyses were used to identify potential variables associated with intraprocedural aneurysm rupture and thrombus formation during coiling. Clinical variables with a p value <0.10 in the univariate analysis were entered into a multivariate logistic regression model. The backward regression method was used to identify independent risk factors for intraprocedural rupture and thrombus formation. Area under receiver operating characteristic curves (AUC) were used to assess the model’s prediction ability and to determine the optimal thresholds of aneurysm size or parent vessel angle that separated rupture and non-rupture groups, or thrombus and non-thrombus groups. An AUC of 0.70–0.79 was regarded as good discrimination and an AUC of 0.80–0.89 was regarded as excellent discrimination. A Kaplan–Meier method was used to depict one minus cumulative survival of patients with thrombus formation dichotomized according to parent vessel angle. ORs and 95% CIs were calculated. A p value <0.05 was considered statistically significant.

Results

Baseline characteristics

The 231 patients included in the study had a mean±SD age of 53.8±11.5 years (range 24–84); 105 (45.5%) patients were female. Intraprocedural rupture occurred in 10 (4.3%) patients, and thrombus formation in 15 (6.5%) patients. Patient characteristics and aneurysm morphologies between intraprocedural rupture and non-rupture groups, and thrombus and non-thrombus groups are presented in tables 1 and 2. One patient had both intraprocedural complications. Rupture occurred in five patients and thrombosis occurred in eight patients between 2008 and 2012, compared with rupture in five patients and thrombosis in seven patients between 2012 and 2015.

Table 1

Univariate analysis of intraprocedural rupture

Table 2

Univariate analysis of intraprocedural thrombus formation

All ruptures were identified during the coil-placement procedure. Intraprocedural rupture occurred in three partially occluded aneurysms and in seven almost completely occluded aneurysms. Eight patients presented with World Federation Neurosurgical Societies (WFNS) grade I and two patients with WFNS grade IV. One patient experienced hydrocephalus and two patients experienced vasospasm after intraprocedural rupture. Of the 15 patients with thrombus formation, four presented with cerebral infarction confirmed by CT scan, and three of the four patients were treated with stent-assisted coiling.

Clinical outcomes at discharge

Clinical outcomes at discharge between groups are presented in tables 1 and 2. Two (20%) patients with intraprocedural rupture died at discharge compared with eight (3.6%) patients in the non-rupture group (p=0.029). Three (20%) patients with thrombus formation died compared with seven (3.2%) in the non-thrombus formation group (p=0.007). Of the 10 patients, five presented with WFNS grade V and seven patients presented with Fisher grade IV at admission. Five patients died of respiratory failure, including two patients with intraprocedural rupture. Three patients died of severe vasospasm, including two patients with thrombus formation. Two patients died of multiple complications.

Predictor of intraprocedural rupture

Patients with smaller aneurysms more often experienced intraprocedural rupture than those with larger aneurysms (3.5±1.3 mm vs 5.7±2.3 mm). Univariate analysis showed that smaller aneurysms (p=0.006), very small aneurysms (≤3.0 mm) (p=0.004), and narrower-neck aneurysms (p=0.016) were associated with intraprocedural rupture (table 1). There was a trend toward intraprocedural rupture in aneurysms with shorter aneurysm domes (p=0.083). There were no statistically significant differences in WFNS grade and Fisher grade between rupture and non-rupture groups. Multivariate analysis showed that smaller aneurysms were independently associated with intraprocedural rupture (p=0.003) (table 3). The threshold of aneurysm size separating rupture and non-rupture groups was 3.5 mm with a sensitivity of 60% and a specificity of 81.9% (AUC=0.80, p<0.001). The receiver operating characteristic curve is shown in online supplementary figure e2.

Table 3

Multivariate analysis of intraprocedural rupture and thrombus formation

Predictor of intraprocedural thrombus formation

Univariate analysis showed that wider-neck aneurysms (p<0.001), smaller dome-to-neck ratios (p=0.047), smaller parent vessel angles (p=0.028), and stent-assisted coiling (p=0.002) were associated with intraprocedural thrombus formation (table 2). There were trends toward thrombus formation in patients with hypertension (p=0.057) and with aneurysms with smaller aspect ratios (p=0.075). There were no statistically significant differences in WFNS grade and Fisher grade between thrombus and non-thrombus groups. Multivariate analysis showed that a history of hypertension (p=0.033), wider-neck aneurysms (p=0.004), smaller parent vessel angles (p=0.023) were independently associated with thrombus formation (table 3). There was a trend toward thromboembolic complication in patients treated with stent-associated coiling (p=0.065). The model including hypertension, aneurysm neck size, and parent vessel angle predicted the risk of thrombus formation with an AUC of 0.84, p<0.001 (table 3).

Parent vessel angle and thrombus formation

The threshold of parent vessel angle separating thrombus and non-thrombus groups was 60.0° with a sensitivity of 73.3% and a specificity of 58.8% (AUC=0.69, 95% CI 0.57 to 0.80, p=0.017); the receiver operating characteristic curve is shown in online supplementary figure e2. If the parent vessel angle was dichotomized into small vessel angle (<60.0°) and large-angle groups, a smaller angle was associated with thrombus formation in the univariate analysis (OR=3.92, 95% CI 1.21 to 12.72, p=0.023). If the small parent vessel angle was included in the initial multivariate analysis, it remained independently associated with thrombus formation (OR=5.68 95% CI 1.48 to 21.77, p=0.011). The dichotomized parent vessel angle in relation to thrombus formation as a function of aneurysm neck size is graphically presented and an illustrative case is shown in figure 1.

Figure 1

Vessel angle and thrombus formation. Kaplan–Meier one minus cumulative survival of patients with thrombus formation according to vessel angle (left). Thrombus formation occurred in a patient with history of hypertension. CT angiography shows maximum aneurysm size (D): 8.0 mm, neck size: 3.8 mm, and vessel angle: 29.6° (right).

Discussion

Using a large number of consecutive patients with ruptured ACoA aneurysms, we found that intraprocedural rupture occurred in 10 (4.3%) patients and thrombus formation occurred in 15 (6.5%) patients. Intraprocedural rupture and thrombus formation were associated with higher risks of hospital mortality than non-rupture and non-thrombus. The threshold of maximum aneurysm size separating intraoperative rupture and non-rupture groups was 3.5 mm, and a smaller parent vessel angle, history of hypertension, and wider-neck aneurysms were independently associated with thrombus formation. A smaller parent vessel angle <60.0° was strongly associated with thrombus formation.

These results are consistent with previous studies showing an intraprocedural rupture rate of 4.1% in ruptured aneurysms,5 ,7 ,8 and a thromboembolic complication rate ranging from 4.7% to 12.5%.3 ,15 ,16 Despite advances in endovascular technology, intraprocedural rupture and thrombus formation remain common during coiling of ruptured ACoA aneurysms, probably because aggressive coiling of ruptured aneurysms to achieve tighter packing increases the risk. Both hemorrhagic and thromboembolic complications were associated with increased hospital mortality. Our results are in line with previous studies that have shown that these complications increase periprocedural death and disability.1–3 However, other studies showed that the complications did not affect clinical outcomes after aggressive treatment.17–19 Differences in patient selection and in methodology might have led to these varied results.

Intraprocedural rupture more often occurs in patients with small ruptured aneurysms.9 ,20 ,21 Small aneurysms are a well-known risk factor for the rupture complication. Mitchell et al22 reported that aneurysms of ≤4 mm were an important risk factor for intraprocedural rupture. However, we found that the threshold of aneurysm size separating rupture and non-rupture groups was 3.5 mm in ruptured ACoA aneurysms. In addition, intraprocedural rupture more often occurred in narrower-neck aneurysms, probably because access to these aneurysms is difficult, and may induce aneurysm perforations.

Our results showed that a history of hypertension, aneurysm neck size, and parent vessel angle were independent predictors of thrombus formation. Hypertension may increase vascular wall tension, vascular inflammation, and remodeling,23 which is a recognized risk factor for stoke. Although wide-neck aneurysms are increasingly treated with stent-assisted and balloon-assisted coiling, these adjunctive techniques may increase the complication rate.3 ,15 Our results also showed a trend toward increased risk of thrombus formation during stent-assisted coiling, probably because of insufficient use of antiplatelet therapy in the early stage of subarachnoid hemorrhage or resistance to antiplatelet therapy.24 Kim et al25 suggested that use of intravenous tirofiban may be an effective approach during stent-assisted coiling of ruptured aneurysms. However, intravenous tirofiban or abciximab before stent placement was not used in our patients because the safety of these drugs was unclear during our study.

Interestingly, we found that a smaller parent vessel angle increased the risk of thromboembolic complication. A possible reason might be that the microcatheter may not remain stable in a smaller parent vessel angle during coiling, and repeat catheterization may induce vessel endothelium injury, platelet aggregation, vasospasm, or clotting. However, a small parent vessel angle does not always hamper the procedure. The underlying reason for the association is uncertain and requires further study.

Limitations

Our study has several limitations. First, it was a retrospective study of patients with ruptured ACoA aneurysms at a single institution. All morphological measurements were made on CTA images, and the limitations of imaging reconstruction with volume averaging may cause subjective measurement bias. The precise measurements require a thorough understanding of postprocessing methods and normal artery anatomy. However, the morphologic parameters were independently measured and their mean values were used for analysis. Second, three-dimensional rotational angiography may provide more reliable information to enable detection of untreated aneurysms.26 However, we did not use rotational angiography images for comparison or internal control, and a comparison of CTA and angiography for the measurement of aneurysm morphologies needs additional study. These results suggest that precise characterization of aneurysm morphology using CTA may help to develop a treatment plan for ruptured ACoA aneurysms. Third, detailed data on procedure phases and duration were not available and could not be used to determine which endovascular devices increased the risks of procedural complications. Finally, these results reflect the experience of a single tertiary referral center and require further validation in additional cohorts. Nevertheless, our results suggested that aneurysm morphologies measured using CTA images may predict the risks of serious complications.

Conclusions

Intraprocedural rupture and thrombus formation remain common during coiling of ruptured ACoA aneurysms. Smaller aneurysms of <3.5 mm were independently associated with intraprocedural rupture. A history of hypertension, wider-neck aneurysms, and a smaller parent vessel angle <60.0° were independently associated with thrombus formation. These high risks of intraprocedural rupture and thrombus formation should be considered before use of endovascular coiling.

References

Footnotes

  • Contributors BZ and YY proposed the study design. LF, BL, TX, NX, and XS were involved in data collection and data verification. BZ, XT, and MZ were involved in the endovascular procedure. LF drafted the manuscript. BZ was involved data analysis and critically corrected the manuscript. BZ, YY, XT, and MZ were involved in data implementation. All authors read and approved the final manuscript.

  • Funding This study was supported by the Chinese Ministry of Health (grant WKJ2010-2-016), the Ministry of Science and Technology of China (grant 2011BAI08B06), and Wenzhou Bureau of Science and Technology (grant Y20140041).

  • Competing interests None declared.

  • Patient consent Obtained.

  • Ethics approval The study was approved by the institutional review board of the First Affiliated Hospital of Wenzhou Medical University.

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