Background and purpose The relationship between dense packing and incidence of angiographic recurrence after endovascular treatment of intracranial aneurysms has been shown but remains controversial. We retrospectively analyzed intracranial aneurysms treated with detachable coils to determine the relation between aneurysm volume, packing, and recurrence.
Methods We reviewed 221 aneurysms in 199 patients who underwent endovascular coiling using detachable coils from November 2009 to December 2011. Aneurysm volumes were determined using three-dimensional images obtained from rotational angiography. Aneurysm packing was defined as the ratio between the volume of coils inserted and the volume of aneurysm. At follow-up, angiographic results were dichotomized into presence or absence of recurrence. The relationship between aneurysm volume to fill, packing, and angiographic recurrence was determined by multivariable logistic regression.
Results Follow-up angiography (mean follow-up 8.8 months) revealed recurrence in 14.5% of the aneurysms studied in our series. Recurrent aneurysms had a mean packing of 15.1% while stable aneurysms (non-recurrent) had a mean packing of 23.7%. Multivariable logistic regression analysis showed that aneurysm volume and packing were significantly associated with angiographic recurrence. Large volume aneurysms (>600 mm3) were found to have a higher incidence of recurrence than those with small volumes (OR=30.49, p<0.001). Compared with those with high packing (≥20%), the less packed aneurysms (<20%) had a higher incidence of recurrence (OR=29.01, p=0.002). There was no significant difference between aneurysm location, clinical presentation, stent assistance, duration of follow-up, and recurrence.
Conclusions Coiling large volume (>600 mm3) intracranial aneurysms are more likely to have a recurrence than small ones. High packing (≥20%) provides better protection against recurrence of the aneurysm.
Statistics from Altmetric.com
Endovascular treatment of intracranial aneurysms with detachable coils has proved to be a safe and effective alternative to surgical clipping1 However, aneurysm recurrence is still a limitation of endovascular coiling compared with surgery.2 Previous studies on large heterogeneous groups of patients with intracranial aneurysms treated by coiling have reported overall recurrence rates of 15–34%.3–5 Factors related to recurrence rates include aneurysms size and neck size,4–6 initial angiographic results,5 ,7 ,8 and rupture status.5 ,9 ,10 However, the relationship between packing and angiographic recurrence remains controversial. Sluzewski et al11 demonstrated that packing plays a critical role in aneurysm recurrence, and that packing of 24% or more was associated with stability of non-thrombosed aneurysms with a volume <600 mm3. On the other hand, Piotin et al12 have shown that packing was not related to aneurysm recurrence. Therefore, the purpose of our study was to retrospectively assess the relationship between packing and angiographic recurrence of aneurysms treated with coils. We also assessed the relationship between recurrence and aneurysm volume, location, clinical presentation, stent assistance, and duration of follow-up.
Materials and methods
From November 2009 to December 2011, 199 patients with 221 aneurysms were enrolled in the present study and treated by endovascular coiling using detachable coils at our institution. Some patients were excluded based on the following criteria: absence of three-dimensional imaging (used for volumetric measurement), no angiographic follow-up, and the aneurysms were partially thrombosed. Our study consisted of 97 men and 102 women, aged 32–86 years (mean 59.6). Of the 221 aneurysms, 37.6% (83) were ruptured and 62.4% (138) were unruptured. Eleven (5%) of the 221 aneurysms were located on the middle cerebral artery, 64 (29%) at the posterior communicating artery, one (0.45%) at the superior hypophyseal artery, six (2.7%) at the anterior choroidal artery, 71 (32.1%) at the ophthalmic artery, 40 (18.1%) at the anterior communicating artery, two (0.91%) at the carotid bifurcation, four (1.8%) at the pericallosal artery, six (2.7%) at the basilar tip, one (0.45%) at the basilar trunk, three (1.4%) at the posterior cerebral artery, three (1.4%) at the posterior inferior cerebellar artery, and nine (4.1%) at the vertebral artery (table 1). All 221 aneurysms underwent angiographic follow-up after a mean interval of 8.8 months (range 1–17).
Volume measurements of aneurysms
Aneurysms volumes were determined using three-dimensional images obtained from rotational angiography (Philips Allua Xper, The Netherlands). The aneurysm was manually segmented from the parent artery, using a volume rending technique, and volume was calculated by the volumetric measurement feature of the system software. Manual outlining and segmentation of the aneurysm from the parent artery was performed by the same physician (BL, who had 15 years of experience in interventional neuroradiology) for all aneurysms. The algorithm was validated with detachable latex balloons (Goldvalve; Nycomed, Paris, France) that were filled with a contrast agent and served as phantoms. The volume filled by the contrast agent (Omnipaque 300; Nycomed) inside the balloons was calculated by weighing the balloon before and after filling. Subtracting both magnitudes resulted in the weight of the contrast agent, and dividing this weight by the density provided the volume of the contrast agent.13 The balloon was imaged with three-dimensional rotational angiography, and volume was calculated using the machine software.13 Errors found during volume calculation using the balloon were approximately 1.6%.
Packing density was defined as the ratio between the volume of the inserted coils and the volume of the aneurysms. Packing was measured as described previously, and calculated in a real time process according to the length of every coil, coil parameters, and aneurysm volumes, which were filled in in the corresponding position of the VB6 program with the graphics interface.14
All procedures were performed with the patient under general anesthesia and access via the transfemoral approach. A 5.0 or 6.0 Fr Envoy guiding catheter (Cordis, Miami Lakers, Florida, USA) was placed in the distal vertebral artery or internal carotid artery. Under fluoroscopic guidance, a microcatheter (Echelon-14; Covidien, Irvine, California, USA) was navigated to the orifice of the aneurysm. Coil placement was performed with detachable coils (Gugliemi detachable coils; Stryker, Neurovascular, Freemont, California, USA; Microplex, Microvention, Tustin, California, USA; NXT fiber coils, Covidien). Generally, a mixture of bare platinum and bioactive coils was used in larger aneurysms while small aneurysms were treated preferentially with bare platinum coils. One or more platinum complex coils were generally deposited in the aneurysm to establish the initial framework. Bare platinum and/or bioactive coils were then deposited to provide additional filling of the aneurysm. Finally, the remainder of the aneurysm was usually filled with small platinum coils. Coils were inserted until the aneurysm was excluded from the circulation or until no more coils could be delivered.
Patients who underwent stent assisted coiling were given antiplatelet agents (75 mg/day clopidogrel and 300 mg/day aspirin) for 3 days before the procedure. These patients received systemic heparinization after placement of the sheath. Activated clotting time was maintained at 2–3 times baseline throughout the procedure. Three types of stents were used to treat intracranial aneurysms, including Neuroform stents (Boston Scientific/Target, Fremont, California, USA), Enterprise stents (Codman, Raynham, Massachusetts, USA), and Solitaire AB stents (ev3, Irvine, California, USA). All stents were deployed following the standard procedure recommended by the manufacturer. The coil microcatheter was positioned into the aneurysm through the stent or before stent placement. Finally, the aneurysm was sequentially coiled using detachable coils.
Angiographic occlusion results were evaluated and classified according to the Raymond Scale, as previously described.5 During follow-up, angiographic recurrence was considered if a previously totally occluded aneurysm had a partial recurrence of the neck and/or sac. In addition, an aneurysm was considered to have remnant growth if a subtotal occluded aneurysm was found to have an increased neck remnant or residual aneurysm.
Statistical analysis was performed using our software (SPSS, V.20). The relationships between recurrence and aneurysm volume, packing, location, clinical presentation, stent assistance, and duration of follow-up were studied with χ2 tests. A p value <0.05 was considered to be statistically significant. The most significant predictors of angiographic recurrence were determined using multivariable logistic regression analysis.
Mean volume for the 221 treated aneurysms was 251.4 mm3 (median 90.9 (range 4–2639)). Mean packing was 22.5% (median 20.4% (range 5.9–57.5%)). The correlation coefficient between aneurysm volume and packing was −0.31 (95% CI −0.038 to −0.233; p<0.001). Follow-up angiography showed recurrence in 32 (14.5%) of the 221 aneurysms at a mean follow-up interval of 8.8 months (median 9 (range 1–17)). In the aneurysms that exhibited recurrence, mean packing was 15.1% (median 15.8% (range 7.5–23.1%)). On the other hand, the aneurysms that remained stable had a mean packing of 23.7% (median 21.3% (range 5.9–57.5%)). Our findings showed that there was a statistically significant difference in packing between recurrent and stable aneurysms.
Aneurysm volume and recurrence
Based on a previously published study,11 we subdivided the 221 aneurysms into three groups: (1) volume <100 mm3, 115/221; volume 100–600 mm3, 79/221; and volume >600 mm3, 27/221. The corresponding spherical diameters for aneurysms with a volume of 100 mm3 and 600 mm3 were 5.8 mm and 10.4 mm, respectively. Recurrence was observed in six (5.2%) of 115 aneurysms with volumes <100 mm3, in 10 (12.7%) of 79 aneurysms with volumes of 100–600 mm3, and in 16 (59.3%) of 27 aneurysms with volumes >600 mm3. There was a significant difference between aneurysm volume and recurrence (p<0.001) (table 2).
Packing and recurrence
Previous studies have demonstrated that achieving a packing of more than 20% was found to protect against recurrence.11 ,15–17 Thus we subdivided our population into two groups. Recurrence was observed in 31 (30.1%) of 103 aneurysms with packing <20% and in one (0.9%) of 118 aneurysms with packing of ≥20%. Aneurysms with packing <20% had a significantly higher risk of developing a recurrence (p<0.001) (table 3)
Location and recurrence
There was no significant relationship between location of the aneurysms and recurrence (p>0.05 for the most frequent aneurysm locations) (table 4).
Clinical presentation (ruptured or unruptured aneurysms) and recurrence
Our data showed that the incidence of recurrence was 18.1% (15/83) in ruptured aneurysms and 12.3% (17/138) in unruptured aneurysms (table 5). Ruptured aneurysms were found to recur more frequently than unruptured aneurysms, but the difference between the two groups was not found to be significant.
Stent assistance and recurrence
The frequency of recurrence was 14.9% (10/67) in aneurysms with stent assisted coiling and 14.3% (22/154) in aneurysms with non-stent assisted coiling (table 6). There was no significant difference between stent assistance and recurrence (p>0.05).
Duration of follow-up and recurrence
Patients were divided into two groups according to the follow-up period: ≤6 months and >6 months. There was no significant difference between duration of follow-up and recurrence (table 7).
Multivariable logistic regression analysis
As can be seen in table 8, aneurysm volume and packing were significantly associated with angiographic recurrence. Aneurysms with large volumes (>600 mm3) had a higher incidence of recurrence than those with small volumes (OR=30.49, 95% CI 7.47 to 158.08). Also, aneurysms with packing <20% had a higher incidence of recurrence compared with those with higher packing (≥20%) (OR=29.10, 95% CI 5.44 to 543.96). There was no significant difference between aneurysm location and recurrence, between clinical presentation and recurrence, between stent assistance and recurrence, or between duration of follow-up and recurrence.
Aneurysm recurrence represents the primary challenge of endovascular treatment compared with surgical clipping, and angiographic recurrences after coil embolization are more common in large or giant aneurysms compared with those of small aneurysms.4 ,5 ,11 Aneurysm volume was previously identified as a significant risk factor. With an increased aneurysm volume, a higher proportion of aneurysms have shown coil compaction. In our study, we subdivided the aneurysms into three groups. We found a significant relationship between aneurysm volume and recurrence; as previously demonstrated, aneurysms with large volumes (>600 mm3) had a higher incidence of recurrence than those with small volumes. Ding et al18 analyzed data on aneurysm volume and histologic healing after coil embolization in a rabbit elastase induced aneurysm model. The authors showed that complete thrombus organization within the aneurysms cavity and endothelium lining the neointima across the neck occurred more frequently in smaller aneurysms than in larger aneurysms. The relatively poorer healing for larger aneurysms can explain high recurrence rates during embolization of larger aneurysms.
Although the relationship between packing and recurrence in treating aneurysms with coils has been extensively studied in the past, it remains controversial as to whether packing plays a critical role in aneurysm recurrence. Some studies suggested that packing of more than 20–25% was found to protect against recurrence.11 ,15–17 Sluzewski et al11 found that if 24% or more of the aneurysm volume was packed, compaction did not occur in aneurysms with a volume <600 mm3. In aneurysms packed more that 20%, compaction did not occur if aneurysm volume was <200 mm3. However, some authors thought that packing was not related to aneurysm recurrence.12 ,19 Goddard et al19 reported an average coil embolization ratio (CER) of 8.2% for 25 small (<7 mm) intracranial aneurysms. The average CER for unchanged aneurysms was 8.0% and for recurrent aneurysms was 8.8%, which was not statistically significant. They concluded that small aneurysms achieved satisfactory stability despite having a low average packing attenuation. CER was not predictive of recurrences in small aneurysms treated with coils. Piotin et al12 reported in their series that recurrent and stable aneurysms both had a mean packing of 27%. They concluded that packing was not related to protection against recurrence. In our study, we subdivided aneurysms into two packing groups. Our results showed that aneurysms with less packing (<20%) had a significantly higher risk of developing a recurrence. Moreover, the relationship between packing and recurrence was also confirmed by logistic regression analysis. These data suggest that higher packing could protect intracranial aneurysms against recurrence.
There are several potential explanations to explain these discrepancies between publications. Firstly, the volumes of aneurysms in our studies were assessed by three-dimensional images obtained from rotational angiography. Three-dimensional rotational angiographic technology has the capability of obtaining precise volumetric measurements.13 ,20 Goddard et al12 measured aneurysm size by comparing with the estimated size of the internal controls, such as the internal carotid artery or the basilar artery. Additionally, they use the formula V=4/3πr3 to calculate aneurysm volume. The erroneous methodology of aneurysm–volume calculation leads to overestimation of the calculation of aneurysm volume and therefore decreasing packing densities. The conclusion that small aneurysms achieved satisfactory stability despite having a low average packing of 8.2% is not totally correct.21 Secondly, undetected intraluminal thrombus might explain why some aneurysms with high packing showed recurrence over time.22 ,23 When initially treated, these aneurysms may contain undetected intraluminal thrombus. While this thrombus resolved, it could result in angiographic luminal enlargement and thus recurrence. Also, high packing may not protect against recurrence by means of thrombus resolution but it does protect against recurrence by means of compaction.
Consistent with previous studies,5 ,9 ,12 our present results showed no significant difference between aneurysm location and recurrence. Some studies published on treatment of intracranial aneurysms in the posterior circulation indicated higher incomplete occlusion, late reopening, and high retreatment rates.24 ,25 This discrepancy might be due to differences in evaluating methodologies and case selection. Considering the surgical risk, large and wide neck posterior circulation aneurysms that have a higher propensity to recur are more frequently referred to endovascular treatment.
The impact of clinical presentation (ruptured status) on recurrence is still unclear. Some studies have shown that recurrence rates were affected by rupture status.5 ,9 ,10 Unruptured aneurysms of <9 mm in size have been reported to have a recurrence rate of 7% whereas ruptured aneurysms of the same size had a higher recurrence rate of 17%.9 Although we and others found that ruptured aneurysms recurred more frequently than unruptured aneurysms, our results showed no significant difference between clinical presentation and recurrence. We further analyzed packing and volumes in these two groups. Ruptured aneurysms had a mean packing of 20.7%±7.9%; unruptured aneurysms had a mean packing of 23.5%±9.7%. Packing of unruptured aneurysms was significantly higher than that of ruptured aneurysms. However, unruptured aneurysms had a mean volume of 292.3 mm3 (median 110, range 4–2639) whereas ruptured aneurysms had a mean volume of 183.4 mm3 (median 81, range 6–2398). The volumes of unruptured aneurysms were larger than those of ruptured aneurysms. We postulate that aneurysm volume might be responsible for blurring what could have been a statistically significant difference.
Stent assisted coiling may reduce aneurysm recurrence after endovascular embolization. Stent placement may provide better angiographic outcomes through improving coil neck coverage or increasing packing density.26 In our study, we found no significant difference between stent assistance and recurrence. A previous study also showed no significant difference in recurrence rates between patients treated with stent assisted coiling and those with non-stent assisted coiling.27 It might be due to achievement of similar packing by using a double microcatheter technique and balloon remodeling in patients without stent placement. We further assessed packing and aneurysm volume between stent assisted and non-stent assisted coiling groups. Aneurysms that were treated with stent assisted coiling had a mean packing of 21.7%±8.8%; those that were treated with non-stent assisted coiling had a mean packing of 22.8%±9.3%. There was no significant difference in packing between the two groups. Aneurysms that were treated with stent assisted coiling had a median volume of 178 mm3 (range 4–2639); those that were treated with non-stent assisted coiling only had a median volume of 71 mm3 (range 6–2477). Aneurysm volume in the stent assisted coiling group was significantly larger than that in the non-stent assisted coiling group (p<0.001). These results might explain why there was no significant difference between stent assistance and recurrence. To evaluate the real impact of stent assistance on aneurysm recurrence, further studies comparing stent assisted coiling with coiling alone for matched aneurysms would be necessary.
The relationship between duration of follow-up and recurrence is also controversial. Previous studies showed that more aneurysm recurrences are detected with longer follow-up.5 ,12 Six month follow-up angiographic study, which is commonly recommended, may be insufficient (too short) to detect all recurrences. In our study, we showed no significant difference between the group followed-up for <6 months and the group followed-up for >6 months. In a previous study of 160 patients, with angiographic follow-up performed at fixed intervals of 6 months and yearly thereafter, almost all recurrences were present at the first 6 month follow-up angiographic examination.28
Our series was biased and might have some limitations with regard of selection of the treated aneurysms. For those patients with ruptured aneurysms accompanied by compressive hematomas, we always prefer surgical clipping. Another limitation is that our study did not exclude the effect of bioactive coils. A mixture of bare platinum and bioactive coils was used during embolization. Despite these limitations, the results of our study demonstrated the relationship between aneurysm volume, packing, and recurrence.
Endovascular treatment with detachable coils is followed by angiographic recurrences in 14.5% of cases. Coiled intracranial aneurysms with a large volume (>600 mm3) are more likely to have a recurrence. High packing (≥20%) seems to provide some protection against recurrence.
BL and YZ contributed equally to this work.
Contributors Conception and design: BL and FX. Analysis and interpretation of the data: BL, YZ, JR, QX, and YT. Drafting the article: FX. Critically revised the article, reviewed final version of the manuscript, and approved it for submission: all authors.
Funding This work was supported by the Advanced Programs of Science and Technology Commission of Shanghai Municipality (09DZ1907503).
Competing interests None.
Ethics approval Ethics approval was provided by Huashan Hospital, Fudan University.
Provenance and peer review Not commissioned; externally peer reviewed.
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.