Single centre experience of stent-assisted coiling of wide-necked basilar tip aneurysms

Abstract

Objective:

To report clinical and radiological follow-up outcomes of stent-assisted coiling of wide-necked basilar tip aneurysms and to evaluate the safety and efficacy of this treatment choice.

Methods:

A retrospective review was carried out of 19 patients with wide-necked basilar tip aneurysms in our institution between 2010 and 2020. The rates of perioperative complication, morbidity, mortality, imaging follow-up and re-treatment were analysed.

Results:

Our technical complication rate was 11% but did not result in treatment failure. The combined procedure related morbidity and mortality rate of the 19 patients who underwent stent assisted coiling was also 11%. 16 of 19 patients had undergone angiographic follow-up with a mean period of 32 months. 81% of patients with angiographic follow-up had a satisfactory occlusion (RROI or II) with 11% requiring re-treatment. Those requiring re-treatment were both treated with laser-cut stents; this is in contrast with no re-treatments required in the patients treated with braided stents.

Conclusion:

Our report provides acceptable outcomes in wide-necked basilar termination aneurysms which are very challenging to treat. Aneurysms treated with braided stents had better efficacy outcomes than those with laser-cut stents.

Advances in knowledge:

Given the emergent and increasing utility of alternative endovascular techniques such as intrasaccular devices and flow diverters, real-world data are lacking on more conventional approaches such stent-assisted coiling, especially so in the posterior circulation. The associated relatively higher aneurysmal haemorrhagic risk in this location warrants further additional safety and efficacy data for this treatment approach, which this paper provides.

Introduction

Since its inception in the late 1990s and early 2000s, stent-assisted coiling has been established as a valid choice, in the management of wide-necked intracranial aneurysms. ^1–4 Compromising 5–8% of all intracranial aneurysms, the basilar tip is the most common site of posterior fossa aneurysms and should be managed more expeditiously, given the reported higher rupture rates and worse post-rupture functional outcomes. ^5–7 Large and wide-necked characteristics to basilar aneurysms further compound these sequelae and present operators with a management challenge. Given the relatively deep location of basilar aneurysms, microsurgical clipping is not deemed to be a first-choice management. Partially thrombosed aneurysms cannot routinely be effectively treated with intrasaccular devices. ⁸ The utility of the stents in the treatment of such aneurysms is to act as a scaffold with the aim of preventing prolapse of coils into the parent vessel, thus enabling a safe degree of packing inside the aneurysm. However successful stent navigation, conformability to the walls of the artery post-deployment, navigation through stent struts with the coiling microcatheter and achieving patency of the parent artery are all challenges which add to the potential for complications.

Several stent-assisted techniques have been developed in the management of wide-necked aneurysms including the deployment of stents in horizontal, T and Y configurations. Other techniques include the utilisation of a single or dual balloons to aid coiling of the aneurysm to achieve a satisfactory level of packing. The challenging nature of treatment of wide-necked aneurysms has resulted in the development of multiple different devices, with varying efficacies. Two such devices, both of which require intrasaccular deployment, the Woven EndoBridge and Contour devices show promising results to date but more long-term outcomes are required before they are firmly established as the standard treatment option. ^9,10 Both of these devices negate the need for long-term antiplatelet therapy, posing an advantage over their stent-assisted alternatives.

The availability of differing types of stents has increased in the past several years including those which are laser-cut and closed cell those which are laser-cut and open cell and those which are braided. ^11,12 It has been hypothesised that the most recent generation of the braided variety of stents, have an increased flow diverting haemodynamic effect, secondary to the higher pore density, when compared with the laser-cut counterparts. ^13,14

This study aims to report clinical and radiological follow-up outcomes of stent-assisted coiling of wide-necked basilar tip aneurysms and aims to evaluate the safety and efficacy of this treatment choice.

Methods

Patient population

We performed a search of our local radiology information system in order to identify patients who had undergone treatment of a basilar tip aneurysm between January 2010 and December 2020. These cases were then individually reviewed in order to identify wide-necked basilar tip aneurysms that were treated with stent assisted coiling. Aneurysms with a neck diameter of greater than 4 mm or a dome-to-neck ratio of less than two were defined as wide-necked. Electronic medical records were reviewed for all patients in addition to selective paper records where electronic record data were inadequate. Radiological records were interrogated by a neuroradiology fellow (HN) and neuroradiology consultant (TP) in order to obtain relevant pre-, intra- and post-procedural data including occlusion scores, aneurysm size and aneurysm location.

Review of clinical records was carried out in order to determine any changes in the patient’s neurological status, using the modified Rankine scale (mRS). Further clinical parameters including presentation type, prior rupture and number and type of prior treatment were recorded. Review of the patient’s post-procedure imaging was carried out to determine the degree of occlusion of the aneurysm, using the Raymond-Roy classification system.

Decisions regarding the most appropriate choice of treatment was collectively determined in our neurovascular multidisciplinary meeting, with representation from both interventional neuroradiologists and neurosurgeons. Multiple factors including complexity of endovascular and open neurosurgical access routes, aneurysm morphology and medical condition of the patient were considered in the decision-making process.

Endovascular procedure

All procedures were performed under general anaesthetic. Systemic anticoagulation was achieved through the administration of an intravenous heparin bolus (at least 3000 IU) and additional top-up boluses administered, guided by the activated coagulation time, in order to achieve an ACT value which was at least 100% greater than the baseline. Arterial access was gained via one (or both in cases requiring two guide catheters) common femoral artery. 6-Fr guiding catheters were placed in one of the vertebral arteries in eight cases, in both vertebral arteries (two catheters) in nine cases and in internal carotid and vertebral arteries (two catheters) in two cases. In the latter of these two cases, a retrograde approach was utilised via the posterior communicating artery (PCom). In select cases, a nimodipine infusion was commenced in order to prevent intraprocedural vasospasm.

A variety of techniques, including T-stenting (Figures 1 and 2), Y-stenting and single stent (horizontal or non-horizontal) assisted coiling (Figure 3), were utilised.

Figure 1.

Incidental partially thrombosed basilar tip aneurysm measuring 6 × 14 x 23 mm. Successful t-stenting using two baby LEO stents (Figure 1a–c). No recurrence at 2-Lyear follow-up imaging (Figure 1d).

Figure 2.

Large wide-necked basilar tip aneurysm incidentally discovered measuring 9 × 11 x 8mm. Successful t-stenting and coiling utilising two baby LEO stents (Figure 2a–b). The aneurysm remains occluded on 6 month and 2-year follow-up catheter cerebral angiograms (Figure 2c–d).

Figure 3.

A patient in their 40s with rupture of a wide-necked basilar tip aneurysm underwent initial emergency coil embolisation (Figure 2a–b). Follow-up catheter angiography demonstrated recurrence at the neck (Figure 2c). This was subsequently treated with a single baby LEO stent (Figure 2d). No recurrence on follow-up imaging.

In the vast majority (95%) of the cases, at least one microcatheter was placed into the daughter vessel for stent delivery (two in cases of t-stenting or y-stenting), whilst another microcatheter was placed into the aneurysm sac. The latter of these was routinely “jailed” before stent deployment. Partial deployment of the stent(s) allowed for the creation of a scaffold to cover the aneurysm neck. Full stent deployment was routinely delayed until a satisfactory coil mass was formed within the aneurysm sac—this was done as a precautionary step to prevent displacement of stent(s) into the aneurysm sac. Subsequently, the stent or stents were deployed, with the placement of additional coils into the aneurysm, via the “jailed” intrasaccular microcatheter, where necessary. A variety braided and laser-cut stents were utilised.

Haemostasis was achieved at the groin access site with the aid of closure devices. The majority of patients were commenced on dual antiplatelet therapy following the procedure for 6–12 months and converted to single-antiplatelet therapy indefinitely.

A technical failure occurring at any stage of the procedure, irrespective of the angiographic outcome, was deemed a technical complication. Clinical complications were also recorded and categorised into periprocedural (within 72 h of the procedure) and delayed (after 72 h of the procedure).

Follow-up evaluation

Angiograms obtained immediately following the end of the procedure were used to assess the occlusion of the aneurysm, according to the Raymond-Roy classification, in addition to assessment of the patency of the stents and interrogation for any vasospasm or thromboembolic complication. Short-term follow-up (carried out within 180 days post-procedure) took the form of mainly MR angiography, with alternatives including CT angiography and digital subtraction angiography. Medium-term follow-up imaging was defined as having taken place between 180 days and 720 days following the procedure and long-term follow-up was defined as having taken place after 720 days post-procedure. Again, these encompassed all of the three aforementioned imaging modalities. The neurological status of all patients was recorded at discharge, and at the very least, at each imaging follow-up, with additional clinic follow-up reviews recorded for selected patients.

Results

Baseline characteristics highlight the complexity of the aneurysms in this series as well as the high number that had been previously treated (Table 1). The endovascular procedure was completed on the first sitting in 18 cases (95%) and in all cases within 24 h of the first sitting.

Table 1.

Baseline characteristics

	Single stent (n = 12)	T-stent (n = 5)	Y-stent (n = 2)	Combined (n = 19)
Male	3 (25%)	1 (20%)	0 (0%)	4 (21%)
Indication for current treatment
Incidental finding	4 (33%)	3 (60%)	0 (0%)	7 (36%)
Subarachnoid haemorrhage	2 (17%)	0 (0%)	0 (0%)	2 (13%)
Recurrence	6 (50%)	2 (40%)	2 (100%)	10 (52%)
Prior treatment	7 (58%)	3 (60%)	2 (100%)	12 (63%)
Prior treatment type
Coiling	6 (85%)	3 (100%)	0 (0%)	9 (70%)
Clipping	1 (15%)	0 (0%)	1 (50%)	2 (20%)
Coiling and clipping	0 (0%)	0 (0%)	1 (50%)	1 (10%)
Mean aneurysm size
Largest diameter	10.6 mm	12.9 mm	9.6 mm	11.1 mm
Neck	5.5 mm	6.7 mm	6.9 mm	5.9 mm
Dome:Neck	1.6	1.1	0.8	1.4

One of the 19 cases required a second sitting for completion of endovascular treatment, 24 h later, following rupture of the right posterior cerebral artery (PCA). This was initially immediately managed with embolisation of the right PCA. A laser-cut stent assisted coiling was performed the following day with the immediate post-procedure angiograms demonstrating a Class 2 occlusion of the aneurysm. The patient’s suffered a right PCA territory infarct with minor residual left-sided weakness on long-term follow-up (mRS 3).

Other than the case of ruptured right PCA, described earlier in this section, there was one further periprocedural complication, bringing the total to 2 (11%). This complication was one of a small dissection of the mid-vertebral artery, causing some compromise to the flow. This was treated with a laser-cut stent with resultant improvement in the flow. This was an asymptomatic complication with the patient discharged with an mRS of 0 following treatment.

There were four deaths in the long-term follow-up in this study, two of which were unrelated to the aneurysm treatment. One patient with a history of two prior treatments of a complex basilar tip aneurysm, underwent a t-stenting approach coiling of a 16 x 8 mm (long axis vs neck) recurrence without any intraoperative complication. Within the short-term post-procedure period, the patient suffered from posterior circulation infarcts highlighted on post-procedure imaging as bilateral thalamic and left occipital infarcts. A root cause analysis revealed unintentional non-administration of antiplatelet therapy for 1 week post-procedure, as the precipitating source. Despite a prolonged period in neuro-rehabilitation, the patient did not recover fully from the resultant neurological deficit and died 9 months later.

A further patient who presented with a rupture from a recurred wide-necked previously treated basilar tip aneurysm, was treated with a single stent-assisted coiling but unfortunately subsequently died in the same admission as a result of her subarachnoid haemorrhage (SAH).

Follow-up evaluation

Sixteen (84%) patients had undergone follow-up angiographic examinations by April 2021 with a mean follow-up length of 32 months. A Class I occlusion was demonstrated in eight patients (50%), with a further five (31%) demonstrating Class II occlusions—none of these patients required further treatment with no reports of rupture on follow-up. Three patients (19%) with recanalization (Class III occlusion) with two of them requiring further treatment in the form of additional coil embolisation.

None of our cases demonstrated a symptomatic in-stent stenosis with none requiring post-procedure angioplasty or re-treatment related to an in-stent stenosis.

At 90-day follow-up, 15 patients (79%) had no deterioration in mRS scores. Two cases of complications related to the procedure include one patient whose mRS changed from 1 to 3 (right PCA rupture) and a further patient whose mRS changed from 0 to 4 (ischaemic due to delay in antiplatelet therapy). Two complications unrelated to the procedure include a further patient with a change of mRS from 0 to 4 (unrelated spinal cord infarct) and the last patient from 5 to 6 (clinically high-grade SAH from pre-procedure).

Discussion

The greatest challenge of endovascular treatment of wide-necked aneurysms, not exclusive to the basilar location, is the prevention of prolapse of coils into the parent vessel and lower complication rates and good long-term efficacy of endovascular treatment. Whilst balloon and stent-assisted techniques have opened several new treatment avenues, endovascular repair remains complex; navigation of stent into daughter vessels, manipulation of microwire/microcatheter systems through stent struts, stent conformability to the artery and preservation of flow within the affected arteries, are to name a few commonly encountered challenges. The latter of these is of paramount importance in the basilar location, given the presence of abundant perforators and potential for debilitating infarcts. The relatively deep location does not favour open surgery and endovascular treatment remains the mainstay treatment for basilar termination aneurysms.

Given that aneurysms treated with these methods were those in which other surgical or endovascular treatment options were not deemed feasible, and in combination with the above-described technical challenges, procedural post-procedural morbidity and mortality in these aneurysms is expected to be higher than their non-complex and non-wide necked counterparts. This is reflected in the rates of recurrence, morbidity and mortality reported in the earlier series describing stent-assisted techniques. ^15–17

In 2004, Chow et al described a novel double stent-assisted technique, with deployment of the stents in a Y-configuration. ¹⁸ Cho et al described a different double stent-assisted technique sometime later, with the deployment of stents in a T-configuration. ¹⁹ They described the stenting technique and presented a small case series of basilar tip aneurysms treated in this manner.

In our study, technical complications occurred in two patients, a vertebral artery dissection and a rupture (and immediately coiled) of the PCA. Unfortunately, this was due to wire manipulation while navigating across the PCA with an intention to do a horizontal stenting across the neck of aneurysm. However, neither resulted in treatment failure and only one resulted in a deficit for the patient, equating to a procedure-related morbidity of 5%. This is in line with the technical complications reported by Bartolini et al and Fargen et al in their respective studies on similar stent-assisted coiling techniques, 6.2 and 6.7%. ^20,21 Although infrequent, the risk of stent dislocation is slightly higher in the Y-stenting technique than T-stenting, owing to the need to navigate through the interstices of the initial stent, to deploy the second, in the former technique.

In our series, the morbidity rate, defined as a deterioration in mRS by Day 90, was observed in four patients, but only two (11%) were related to their procedure—one patient who was not administered antiplatelets due to human error and another patient who suffered a PCA rupture managed with PCA sacrifice which resulted in an expected infarct and resultant neurological deficit. The former of the aforementioned patients succumbed to their morbidities resulting in a procedure related mortality rate of 5% for our series which likely reflects the inclusion of emergency cases in our series (11%). Additional factors that may be contributory to mortality and morbidity include the potential for increased overall haemorrhagic risk (owing to the need for antiplatelet therapy) as well as and increased potential for thromboembolic complications dependent on patient responses to antiplatelet therapy. The vast majority of patients included in this series underwent their procedures at a time when platelet function testing was not available at our institution, meaning that antiplatelet treatment strategy was not individualised as it is today. There is a wide range in the reported rates of complication in the literature, with studies reporting as low as 0% to as much as 42.1%, which likely reflects the differing patient populations, intricate stenting techniques and challenging aneurysmal anatomy. ^22–24 It is worth mentioning that the single case of procedure-related mortality in our series, was an unfortunate incident given that the human error of missing antiplatelet administration is an avoidable one.

By minimising the risk of coil prolapse into parent vessels, the use of stents allows a more assertive aneurysm packing strategy, allowing higher rates of aneurysm occlusion. This is reflected in our occlusion rates of 81% (RRO I or II) in patients with a minimum of 4 month follow-up. These rates are in line with both larger- and similar-sized studies respectively carried out by Chalouhi et al (86%) and Zhang et al (75%). ^25,26 Of the three patients who had RRO III on their follow-up imaging, two underwent re-treatments with further coiling procedures putting our re-treatment rate at 11%. Although this is slightly higher than Chalouhi et al’s study (7.8%), it is considerably less than that of the higher re-treatment rates reported in non-stent assisted coiling of basilar aneurysms. ^25,27

One of the above-mentioned patients had a laser-cut stent in a horizontal orientation, with access gained into the P1 segment of the right PCA via the right PCom and the stent placed across the basilar tip aneurysm extending into the P1 segment of the left PCA. The jailed coiling microcatheter was subsequently used to coil the aneurysm. Follow-up imaging at 2 years demonstrated coil compaction with a significant recurrence requiring a further coil embolisation. Additional follow-up imaging demonstrated further recurrence at 5 years following the initial procedure which was coiled for a third occasion. Follow-up imaging at 7 years showed stability with no significant recurrence.

The above-described patient, in addition to the only other retreated patient in our series (Table 2) has led us to ponder on whether contributing factors to significant recurrences in the stent-assisted treatment of wide-necked basilar tip aneurysms includes the choice of stent—i.e. braided vs laser cut. Whilst our case series is small, the rates of satisfactory occlusion (RROC I or II) in the braided stents (100%) versus laser-cut stents (67%) cannot be ignored. In their systematic review of 35 series, Cagnazzo et al highlighted the high (88–89%) rates of total or near-total occlusion with a couple of braided stents and suggested that these could be due to the higher metal coverage, smaller cell size and flow diverting properties of such stents. ²⁸ Drawing a direct comparison, Xue et al demonstrated a recurrence rates favouring aneurysms treated with a braided stent, when compared with a laser-cut one (3.7% vs  9.6%, total of 235 patients). ²⁹ It is however worth noting that in the only case of recurrence which was not retreated in our series, a case with use of a laser-cut stent, this decision was taken due to stabilisation of the recurrence, which has been closely monitored for over 72 months. It is important to bear in mind that treatment options in these cases tend to be complex and as such the first step should not always be to re-intervene; a conservative more regular imaging surveillance strategy remains a valid option in select cases.

Table 2.

Radiological outcomes in 16 of 19 patients with follow up

	Mean follow-up length	Raymond Roy occlusion grade			Re-treatment rate
		I	II	II
Stent type
Braided (7)a	24 months	71% (5)	29% (2)	0% (0)	0% (0)
Lasercut (9)b	39 months	33% (3)	33% (3)	33% (3)	22% (2)
Technique
T-stent (4)	25 months	50% (2)	50% (2)	0% (0)	0% (0)
Y-stent (2)	31 months	100% (2)	0% (0)	0% (0)	0% (0)
Single-stent (10)	36 months	40% (4)	30% (3)	30% (3)	20% (2)
All cases
(16)	32 months	50% (8)	31% (5)	19% (3)	13% (2)

^aAneurysms treated with braided stents—(6 x Baby Leo, 1 x Baby Leo and LVIS EVO).

^bAneurysms treated with laser-cut stents—(3 x Enterprise, 2 x Solitaire, 2 x Barrel, 1 x Acclino, 1 x Enterprise and Solitaire).

Additionally, it would be prudent to not take our results as indicative of an end to the utility of laser-cut stents, as we have shown that they too, can provide successful outcomes in aneurysms requiring a highly complex treatment. Interventionists ought to be aware of potential differing outcomes and continue to make their selections on stent choice on a case-by-case basis taking into account many factors including aneurysm and target vessel morphology amongst others.

Limitations of our study includes the relatively small size and retrospective non-randomised nature of our study. The lack of a control group undergoing alternative endovascular treatments and selection bias are further recognised limitations.

Conclusion

In this report, we assessed clinical and radiological outcomes of stent assisted coiling in wide-necked basilar tip aneurysms at our centre. Our results demonstrated acceptable outcomes in wide-necked basilar aneurysms which are very challenging to treat. Aneurysms treated with braided stents had better efficacy outcomes than those with laser-cut stents. The stent assisted treatment comes at a higher, but acceptable rate of complication with the trade-off being one of higher efficacy.