Endovascular reconstruction of aneurysms with a complex geometry
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.178034
Source of Support: None, Conflict of Interest: None
Conventional endovascular coiling remains the mainstay of treatment for most aneurysms; however, it may not be suitable for aneurysms with a complex geometry and there remains the risk of recanalization. Aneurysms with an unfavorable morphology are difficult to treat through both endovascular and surgical means. Progress in endovascular technology has allowed for the emergence of newer strategies to treat aneurysms with a complex geometry. Better packing density in wide-necked and large aneurysms can be achieved through the balloon remodeling technique. Similarly, a self-expanding stent cannot only act as a scaffold that helps to retain coils but also aids in diverting the blood flow away from the aneurysm sac. Lately, focus has shifted from endosaccular occlusion to endoluminal reconstruction; flow diverters are being increasingly used to treat aneurysms with an unfavorable geometry. However, there is no clear consensus on the best endovascular management strategy in certain subset of aneurysms - large and giant internal carotid aneurysms, blister aneurysms, and fusiform/dissecting aneurysms of the vertebrobasilar artery. We present a review of literature and discuss the current evidence for the various endovascular strategies to treat complex aneurysms.
Keywords: Assisted coiling; complex aneurysms; endovascular reconstruction; flow diversion; geometry
Endovascular management of cerebral aneurysms has progressed with time, and the International Subarachnoid Aneurysm Trial suggested that the morbidity and mortality may be lower with endovascular intervention as opposed to surgical treatment in ruptured aneurysms.  Consequently, endovascular coiling has become the preferred method to treat intracranial aneurysms at many centers. However, recanalization of the aneurysmal sac remains an important concern. Certain aneurysmal morphologies particularly pose a specific technical challenge to both the endovascular and surgical treatment strategies. To address these issues, endovascular management has evolved from the deconstructive techniques and conventional coiling to adjunctive coiling (balloon-assisted), and more recently, reconstructive techniques (stent-assisted and flow diversion). Conventional coiling remains the mainstay of treatment in most aneurysms; however, certain sac characteristics herald the need for balloon or stent assistance. The balloon remodeling technique involves temporary inflation of a nondetachable balloon across the neck of the aneurysm sac to allow for coil placement and better packing, particularly in wide-necked and large aneurysms. Similarly, a self-expanding stent can prove to be an effective scaffold that helps retain the coils in place. In addition to providing support, they act as conduits that divert flow from the aneurysmal sac and allow for neointimal growth on the inner wall resulting in healing of the vessel wall.
Flow diverters are being increasingly used to treat aneurysms with an unfavorable sac geometry; the key underlying principle being endoluminal reconstruction. A flow diverter stent has a high pore density, low porosity, and approximately 30-35% metal coverage on the inner surface of the parent artery. This allows for flow redirection into the parent artery thereby promoting sac thrombosis, neointimal proliferation, and reconstruction of the parent artery.  Better sac occlusion resulting in a low rate of recurrence was noted in wide-necked large and giant aneurysms treated by this technique. We present a review of the current status of endovascular reconstruction of complex aneurysms with an unfavorable sac geometry (large and giant anterior circulation aneurysms, blister aneurysms, and posterior circulation fusiform aneurysms). Bifurcation aneurysms have not been included due to the limitations of the length of the article permissible.
Cavernous carotid large and giant aneurysms are difficult to manage and various endovascular techniques have been proposed to treat these aneurysms; they include coil embolization with or without adjunct, parent artery occlusion (PAO), onyx embolization, and lately flow diversion.
Coil embolization of large and giant aneurysms is safe but not without drawbacks. In general, aneurysms packed to approximately 25% or more of its luminal volume remain stable at follow-up.  However, in an in-vitro model, a packing density of 9% is at best achieved with large and giant aneurysms when the sac appears well packed. Therefore, there remains the risk of coil compaction and significant recanalization.  An aneurysm occlusion rate with coiling was approximately 67% on interval imaging according to Turfe et al. 
The safety of PAO has been evaluated in multiple studies. An early systematic review suggested higher rates of ischemic permanent complications with balloon PAO (5%) as compared to coiling (0%); however, the venous phase delay was assessed only in 26.5% of the patients.  Later, it was determined that venous phase delay on balloon test occlusion (BTO) was a crucial surrogate marker for adequacy of perfusion in the occluded territory.  They noted no morbidity or mortality with PAO when the venous phase delay was ≤2 s. Similar good outcome with PAO was noted by van Rooij.  On PAO, when performed after a successful BTO, near complete or complete resolution of symptoms due to mass effect, was noted in 92% of patients without any direct procedure-related permanent neurological morbidity. Furthermore, Ishii et al., (2014) reported on the safety of PAO based on a review of patients in the Japanese registry for neuroendovascular therapy. Major/minor stroke or transient ischemia was noted in 5.8% of patients treated with PAO after successful BTO. Permanent morbidity and mortality at 30 days was noted in 3.1% and 1% of patients, respectively.  Contrary to the findings of this study, a recent systematic review comparing endovascular coiling (with or without stent assistance) with PAO (without bypass) suggested a tendency toward greater morbidity and mortality with PAO.  Nonetheless, this difference was not statistically significant and the principal criticism against the study was the lack of mention of the proportion of patients who had venous phase delay assessment on BTO, the key criterion used to determine the safety of the procedure. In our experience, a short-term antiplatelet and anticoagulant administration along with maintenance of cerebral perfusion pressure are crucial to avoiding cerebral ischemia.
The possibility of development of de novo aneurysms has been raised as an issue after PAO. Although Arambepola et al., noted a tendency for the formation of new aneurysms in this patient group, this was not confirmed in other studies. ,, Therefore, there is no definite evidence to support this concern and the risk of developing de novo aneurysms should not prevent the physician from offering this therapy to the patient.
The concept of reconstructing the parent artery appears to be a promising option particularly for large and giant cavernous carotid aneurysms (CCAs). Puffer et al., reported on near complete or complete aneurysm occlusion in 83%, and complete resolution of symptoms in 90% of patients, treated with a flow diverter stent. Remarkably, there was no procedure-related morbidity or mortality.  Similar results were observed in CCA by O'Kelly et al.  Twenty-six CCA were treated with a pipeline embolization device (PED); the mean diameter was 21.4 mm, and 38% had partial thrombosis of the fundus. No intracerebral hemorrhage, aneurysmal rupture, or major thromboembolism was noted. Sixty-one percent of patients presenting with cranial nerve deficits had complete resolution of their symptoms. Therfore, flow diversion has emerged as a safe alternative strategy to treat large and giant aneurysms of the CCA; this is particularly true in the group who are intolerant to BTO.
Certain conclusions can be inferred from the results of the mentioned studies. Recanalization rates are high, needing a repeat treatment in the group managed with coiling (with or without adjunct); and, therefore, coiling is not the best strategy to treat a CCA. The PAO in patients with a venous phase delay of ≤2 s is an effective and safe strategy to treat large and giant CCAs. Resolution of the mass effect and aneurysm occlusion rates are significantly better than that seen when coiling alone was done. Thromboembolism secondary to an embolic etiology is not infrequent and therefore, administering antiplatelet agents prior to and for a short-term following the procedure can be considered. Caution should be exercised while interpreting the results of the systematic reviews as in the majority of the studies, the venous phase delay was not assessed prior to the PAO. Flow diversion appears to be an effective strategy to treat these aneurysms. Short-term results with flow diversion appear to be positive and indicate that this procedure is an effective choice particularly when the venous phase delay is >2 s.
Large and giant aneurysms are not uncommon and approximately 17% of aneurysms were ≥15 mm in a large series reported by van Rooij and Sluzewski.  These aneurysms are prone to rupture; the annual rupture rate was approximately 4.7% for aneurysms measuring 10-24 mm, and 33.4% for aneurysms >24 mm in a Japanese cohort.  The morphology, however, imposes certain limitations on the routine surgical and endovascular treatment methods. The prime benefit obtained with the surgical management is a better occlusion rate and lack of mass effect; however, the combined surgical morbidity-mortality ranges from 16.9% to 34%. ,, In addition, the International Study of Unruptured Intracranial Aneurysms investigators noted that poor outcome was common with surgical intervention as compared to endovascular management of large and giant anterior circulation aneurysms. 
Many endovascular strategies have been proposed to treat these aneurysms and include the following: Coiling, adjunctive coiling, and embolization using liquid embolics. The conventional coiling is a safe strategy to treat these aneurysms; however, the key drawback is the high recanalization rate resulting in the need for a repeat treatment. Retreatment rate of over 50% has been reported in the literature. , To achieve better packing and reduce the risk of recanalization, both balloon-assisted coiling (BAC) and stent-assisted coiling (SAC) have been considered as acceptable alternative strategies. However, among the adjunctive coiling techniques, better occlusion rates were achieved with SAC as opposed to BAC. ,
In addition, stent assistance, when used as a scaffolding to treat large and giant aneurysms, did not predispose to higher complication rates.  Particularly, the retreatment rate was a third less likely with SAC (26%) as opposed to conventional coiling (37%). Similarly, new onset bleeding or the rebleeding rate was relatively low with SAC (1.3%). Therefore, SAC appears to be an effective strategy to treat large and giant aneurysms. Nevertheless, aneurysm occlusion rates appear to be less than optimal even with stent assistance.
Hypothetically, endoluminal reconstruction that allows for aneurysm exclusion by way of sac thrombosis and neointimal proliferation appears to offer the best chance to treat these aneurysms. Therefore, there is a growing enthusiasm to use flow diverters in this aneurysm category. In a large systematic review of predominantly unruptured, anterior circulation, large and giant (50%) aneurysms treated with the pipeline embolization device (PED), the cumulative mortality rate was 2.3%. The early and delayed complication rates were 5.7% and 1.9%, respectively. Near complete or complete occlusion was noted in 80% of the aneurysms at a 6 month follow-up. In addition, the need for retreatment is low with flow diversion.  Therefore, the complication rates are comparable with, and aneurysm occlusion rates are better to that seen in SAC. In addition, even though a delayed clinical deterioration due to a transient increase in the perianeurysmal brain inflammation can occur following the flow diverter placement, complete resolution of the mass effect was noted in most patients with anterior circulation aneurysms. Aneurysms with a large diameter (mean 2.2 cm), and the aneurysmal sac embedded in the brain parenchyma without intervening cerebrospinal fluid, are predictors of peri-aneurysmal brain inflammation. ,, In ruptured aneurysms, conventional coiling can be done in the acute stage to prevent rebleeding from the aneurysm. This may be followed by the flow diverter placement at a suitable time interval to complete the neck reconstruction [case study 1 illustrated in [Figure 1].
In conclusion, conventional coiling, although safe, is fraught with higher recanalization and bleeding rates. Among adjunctive coiling techniques, occlusion rates are better with SAC in comparison to BAC. However, even with SAC, the occlusion rates are less than expected and in approximately one-fourth of the patients, retreatment is required. Finally, a better occlusion of the aneurysm resulting in lower retreatment rates with the flow diversion procedure makes it a very effective strategy to treat these large and giant aneurysms in the anterior circulation and may be the treatment of choice in carefully selected patients.
Blister aneurysms (BAs) differ pathologically from saccular aneurysms in that they lack intima and media; they are small with blood blister-like morphology, tend to grow without treatment, and have a predilection for the dorsal supraclinoid internal carotid artery (ICA). Diagnosis of blister aneurysms (BA) can be easily overlooked on computed tomography angiography, and for that matter, even on two-dimensional digital subtraction angiography. However, with the advent of 3D rotation angiography, aneurysms <3 mm can be accurately detected. 
Both surgical and endovascular management of these aneurysms can be extremely challenging. In a systematic review, the overall combined morbidity and mortality with surgical management was reported to be as high as 32%.  With respect to endovascular management, these aneurysms are small, highly friable, and broad-necked and therefore, present the risk of sac rupture during catheterization and coiling. Thus, stent assistance is needed to retain the coils in place. Various modalities have been suggested: Coiling, SAC, stent (single/overlapping) alone, PAO by trapping, and flow diversion. The concern, however, is the requirement of an antiplatelet agent in a partially secured aneurysm in the setting of acute subarachnoid hemorrhage (SAH) if stent-assisted treatment is contemplated.
Rebleed and regrowth of the aneurysm necessitating a repeat treatment was noted in 12.5% and 34.5% of ruptured blister aneurysms (n = 87) managed by the endovascular intervention.  The combined morbidity and mortality rate was 10% with SAC, and 8% with overlapping stents. In another series of predominantly ruptured blister aneurysms treated with SAC, a good functional outcome was noted in 92.3% of patients.  Rebleed resulting in death was noted in 1 (7.6%) patient and a complementary interval treatment was required in 2 (15.3%) patients. Furthermore, our group reported on 17 ruptured blister aneurysms that were managed by endovascular intervention. Mortality due to rebleed occurred in one patient (5.8%) and a repeat treatment for regrowth was required in two (11.6%) patients.  In both the above series, the rebleed and regrowth rates were lower than that reported in the systematic review and could be attributed to the variability in the modality of endovascular treatment offered to the patients.
Flow diversion is now being increasingly used to treat BAs. Conceptually, flow redirection is better with flow diverters as compared to the traditional stents. A flow diverter also leads to intimal growth along the stent allowing for the possibility of "healing" of the aneurysm. Aydin et al., noted no rebleeding or thromboembolic complications, and a complete aneurysm sac occlusion was achieved in all patients treated with flow diversion for a ruptured BA.  However, they delayed the deployment of the device until the fifth day to avoid the early administration of antiplatelet agents. Similar aneurysm occlusion rates and a good outcome was observed in the seven patients reported by Çinar et al.  At our center, we have treated 9 ruptured blister aneurysms (unpublished data) in the acute setting [case study 2 illustrated in [Figure 2]. Treatment was not delayed and the patients were administered antiplatelet agents 2-h before the deployment of the stent. The need for an external ventricular drain (EVD) was assessed and if required, it was performed before administering antiplatelet agents. There were no incidence of rebleeding and almost complete sac occlusion was confirmed on an early interval angiography in all the treated patients.
It appears that SAC and overlapping stent placement are suitable treatment modalities for BA; the combined morbidity-mortality is significantly less than that seen with surgical treatment. However, with single/overlapping stent strategy, there is the possibility of rebleed and regrowth. Recently published series and our own experience indicate that the flow diverter placement may emerge as the treatment of choice in these patients.
Aneurysms with a fusiform morphology can represent either a chronic fusiform aneurysm or an acute dissection with ectasia. Dissecting aneurysms frequently present with SAH or acute ischemia. Both surgical and endovascular management of these aneurysms can be challenging. Surgical morbidity is however high with the mortality rates as high as 25% seen with surgical PAO.  We have discussed the vertebral artery (VA) and basilar artery aneurysms (BAA) management separately as the choice of endovascular techniques can differ greatly between the two groups.
Ruptured VA dissecting aneurysm if left untreated carries a poor prognosis; the tendency to grow and rebleed ranged from 30% to 70%. , Both deconstructive and reconstructive methods have been used to treat these aneurysms.
Proximal occlusion with flow reversal and trapping are the two techniques that have been described. The above techniques should be considered in patients with a good caliber contralateral VA. Aneurysm occlusion was noted on the early interval angiography in all the patients treated with trapping; whereas, persistent sac filling was noted in the proximal occlusion cohort.  Furthermore, rebleeding and a poor outcome were frequent in the patients treated by proximal occlusion as compared to trapping. , Therefore, whenever possible, trapping should be the preferred treatment option for VA dissecting aneurysm.
Parent artery aneurysmal segment reconstruction is a valuable option particularly in patients with a hypoplastic or occluded contralateral VA and a poor Willisian collateral circulation (mainly the posterior communicating artery). According to Sönmez et al., although early and long-term sac occlusion was better with trapping, the perioperative morbidity and mortality, recurrence and retreatment rates, and long-term clinical outcome were similar with both the deconstructive and reconstructive techniques.  The reconstructive techniques include SAC, a single stent, overlapping stents, and flow diversion. With SAC, we advocate using multiple small coils to pack the sac in order to avoid extraneous outward pressure on the wall. When residual filling is noted, overlapping stents could be considered. There are fewer perforators in the intradural VA than in the basilar trunk; therefore, flow diversion appears to be an effective strategy to treat these aneurysms. A good outcome was noted in all but two patients (83%) when predominantly symptomatic aneurysms were treated with flow diversion.  A similar good outcome and complete sac occlusion were also noted with flow diversion in asymptomatic VA aneurysms. ,,
In conclusion, trapping results in better immediate and long-term sac occlusion rates and therefore, appears to be the best technique to treat these aneurysms. Contrary to this, the risk of recanalization and retreatment exists with the reconstructive techniques (SAC and overlapping stents). The latter procedures should, therefore, be reserved for patients with an occluded or hypoplastic contralateral VA and poor Willisian collaterals. The preliminary studies show that flow diverters can be considered as an acceptable alternate strategy; however, this needs further evaluation in larger studies.
Basilar trunk fusiform aneurysms (BTA), when symptomatic, have a poor prognosis; progressive brain stem compression or SAH can occur with the 2-year survival rate being as low as 20%.  They are difficult to manage and unlike in VA aneurysms, PAO is not an option in most patients with BTA. The fusiform morphology heralds the need for stent assistance and in turn antiplatelet administration. However, one cannot be certain of complete aneurysmal exclusion and, therefore, the risk of rebleed. Furthermore, the basilar trunk is rich with perforators; the presence of a large sac and a mural thrombus before the starting of treatment can, in effect, precipitate the formation of sudden and significant clot burden resulting in occlusion of the perforators. Finally, the age of the patient, the size of the aneurysm, and the symptomatic status of the aneurysm (presence of mass effect or rupture) are the key predictors of outcome following treatment. The reconstructive techniques used to treat BTA are as follows.
The stent type, number, and the method of deployment appear to determine the outcome in this group of patients. Li et al., reported on 11 symptomatic BTA treated predominantly with laser cut stents; the stents were deployed in a truly overlapping fashion only in a small proportion of patients.  In 50% patients, no change in the sac filling was noted and 27% had a poor outcome due to a posterior circulation stroke.
Contrary to these results, Kizilkilic et al., reported on satisfactory aneurysm sac occlusion without in-stent stenosis and a good outcome in all the four patients treated with overlapping braided stents. An EVD and an occipital craniectomy along with C1 vertebra posterior arch removal were performed prior to the definitive endovascular treatment for patients presenting with mass effect. Therefore, it appears that braided stents deployed in a telescopic manner across the aneurysm sac may offer better flow diversion and aneurysm sac occlusion in comparison to a single stent. 
Stent and coil
Partial or complete aneurysmal sac occlusion was noted in 92% of patients (n = 13) treated predominantly with stent and coil.  Coils were deployed in 92% of patients, and a braided stent was used in most. The second treatment included placement of a flow diverter and occlusion of a single VA to reduce flow. Poor outcome was noted in 23% of patients; two had ischemic complications, and the aneurysm in one of the treated patients rebled after a month.
Basilar trunk is rich with perforators; therefore, there is a reluctance to use flow diverters. A poor outcome was noted in 66% of patients treated with a SILK flow diverter for BTA.  A similar poor outcome was noted in all patients with a giant fusiform BTA predominantly due to brain stem ischemia.  Contrary to the above, all seven BTAs treated with a PED had a good outcome.  Similarly, sac occlusion was complete in 83%, and a good outcome was seen in 66% of patients treated for a symptomatic large or giant BTA.  The wide variability in the outcomes achieved with flow diversion can be explained by the heterogeneity in the aneurysm sac and patient characteristics included for treatment.
Basilar trunk fusiform aneurysms, particularly the large and giant aneurysms, are difficult to treat; therapeutic modalities should be individualized based on the sac characteristics. Most studies presented above are limited by the small number of patients recruited for the study, and therefore, definite conclusion cannot be drawn from them. A braided stent, by achieving better wall opposition by adapting to the vessel anatomy and a higher metal coverage, may offer better flow diversion in comparison to the laser cut stents. Overlapping braided stents with or without coil deployment may allow for a better aneurysm occlusion and a good outcome [case study 3 illustrated in [Figure 3]. Occlusion of a single VA can be considered to further reduce flow into the aneurysm sac over and above the placement of the stent and coil. Flow diverters are suitable alternatives in carefully selected patients with a fusiform BTA.
Specific endovascular reconstructive techniques have been developed to address the challenges faced while treating an aneurysm sac with a complex geometry. In suitable patients, PAO, a deconstructive technique, can be considered as an initial treatment option in cavernous carotid large and giant aneurysms, and in symptomatic fusiform aneurysms of the VA. Among reconstructive techniques, flow diversion seems to be an effective strategy to treat large and giant, as well as blister aneurysms of the ICA. A stent with coil, a stent (single/overlapping) alone, and flow diversion are all options for vertebrobasilar dissecting aneurysms. The treatment should be individualized according to the sac morphology.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3]