Abstract
Background
Both low-profile visualized intraluminal support (LVIS)stents and Pipeline flow diverters (FDs) are therapeutic strategies for basilar artery (BA) aneurysms, but they have not been directly compared.
Methods
A total of 132 consecutive patients with 139 BA aneurysms treated with either LVIS stent or Pipeline FDs were analyzed. Propensity score matching (PSM) was used to control for age, sex, hypertension, aneurysm size, shape, location, and duration of follow-up. The treatment results of these two braided stents were compared.
Results
LVIS stent was placed in 88 (63.3%) and Pipeline FDs in 51 (36.7%) procedures. Patients with Pipeline FDs tended to be younger and have less hypertension, whereas aneurysms had larger aneurysm size. After PSM, similar complete or near-complete occlusion rates (76.7% vs 73.3%, p = .766) and favorable functional outcomes (86.7% vs 90.0%, p = 1) were achieved in patients treated with LVIS stents and Pipeline FDs, respectively. Further comparisons were conducted at three different locations (basilar apex/basilar trunk/vertebrobasilar artery junction [VBJ]) separately, and the results showed a higher complete or near-complete aneurysm occlusion rate after Pipeline FD treatment than LVIS treatment (86.7% vs 59.2%, p = .012) only at VBJ, where a particularly high proportion of non-saccular shape (70.9%) and a male preponderance were noted.
Conclusion
Both braided stents were effective in the treatment of BA aneurysms, with good occlusion rates and favorable functional outcomes. Pipeline FD achieved a particularly higher aneurysm occlusion rate than LVIS stent at VBJ, where lesions often require reconstruction of the diseased vessel.
Keywords: Intracranial aneurysm, stent, basilar artery, treatment
Introduction
Intracranial aneurysms are found in approximately 3% to 5% of the population, and the annual rate of rupture is approximately 10 per 100,000 individuals. Posterior circulation aneurysms constitute approximately 15% of all intracranial aneurysms and can be saccular or non-saccular (fusiform/dissecting) shapes. 1 Basilar artery (BA) aneurysms account for 2% to 3% of intracranial aneurysms, including the basilar apex, basilar trunk, and vertebrobasilar junction (VBJ) aneurysms.2,3 The treatment strategy for BA aneurysms remains one of the most challenging and controversial topics of cerebrovascular neurosurgery. Due to the deep location of the BA aneurysms and their supply to many critical structures, the microsurgical approaches for the treatment of BA aneurysms often require significant brain retraction and temporary arterial occlusion, which may result in significant morbidity and mortality. On the other hand, the typically straight course and large caliber of the BA aneurysms make it particularly favorable to an endovascular approach and endovascular treatment has evolved into the mainstay in treating these lesions.2,3 However, the safety and efficacy of endovascular treatment for BA aneurysms have not yet been established definitively. Given the relatively small number, BA aneurysms are a diverse group and include varying morphologies and anatomical configurations.2,3 Large and non-saccular aneurysms are frequent, posing increased technical challenges to traditional endovascular techniques. Braided intracranial stents, including flow diverters (FDs) and low-profile braided intracranial stents, may provide additional advantages due to their flow-diverting properties. 4
Tightly braided Pipeline FDs (Pipeline embolization device; Covidien/Medtronic, CA, USA) have low porosity and provide 30%–35% metal coverage, and the aim is to reconstruct the parent vessel, reduce intra-aneurysmal flow, and finally occlude the aneurysm after subsequent thrombosis and cervical neointimal coverage. In recent years, FDs have been increasingly used for the treatment of posterior circulation aneurysms. However, a recent long-term multicenter experience study by Adeeb et al. 5 reported a trend toward treating fewer BA aneurysms but more vertebral artery aneurysms with FDs across the years. BA aneurysms with FDs were identified as independent predictors of thromboembolic complications. Interventionalists have raised concerns about the impact on the abundant perforators along the BA. 5 On the other hand, low-profile braided stents (Low-profile Visualized Intraluminal Support-LVIS; MicroVention, CA, USA) have a similar braided design 6 but less thick mesh (17%–23% metal coverage) than FDs. They were destined for stent-assisted coiling, as a scaffold or support for coils for especially anatomically challenging complex aneurysms. Their relatively less-dense braided mesh theoretically could lead to some extent flow-diverting properties, while allowing for easier navigation through thin microcatheters and less metal coverage in front of perforators. They have shown lower complication rates in distal intracranial locations than FDs, while still maintaining high and stable aneurysm occlusion rates. 6
Here, we report a 7-year consecutive experience in treating location-specific (i.e., BA only) aneurysms in two medical centers, with a larger sample size of 139 and a focus on the comparison of two braided stents (LVIS and Pipeline) for treating these BA aneurysms. Propensity score matching (PSM) will help to reduce confounding biases.
Materials and methods
This retrospective study was approved by the local Institutional Review Board. From January 2015 to June 2022, a total of 132 consecutive patients with 139 untreated BA aneurysms who received either LVIS stents or Pipeline FD treatment in our two medical centers were enrolled. They were divided into the LVIS group and the Pipeline group and compared. Patient baseline data (age, sex, alcohol drinking, smoking habit, comorbidities, and with multiple aneurysms), aneurysm characteristics (location, rupture status, shape, and size), procedural details (type of stents, stent numbers, and adjunctive coiling), and radiographic and clinical outcomes were recorded (Tables 1 and 2). When the same patient with multiple aneurysms received multiple braided stent treatments, they were considered different aneurysms. The shape was defined as saccular or non-saccular. The size included the maximum diameter for a saccular-shaped aneurysm or the maximum length for a non-saccular-shaped aneurysm. DSA follow-up was scheduled at 3, 6, and 12 months after endovascular treatment. Patients with completely occluded aneurysm do not need further DSA follow-up but receive MRA or CTA follow-up annually.
Table 1.
Patients’ baseline characteristics.
| Variable | Patients with braided stents | p-Value a | ||
|---|---|---|---|---|
| Total (n = 132) | LVIS (n = 85) | Pipeline (n = 47) | ||
| Age, years | 50.3 ± 13.9 | 53.3 ± 10.3 | 44.8 ± 17.7 | .001 |
| Female (%) | 82 (62.1%) | 52 (61.2%) | 30 (63.8%) | .91 |
| Smoking habit (%) | 30 (22.7%) | 19 (22.4%) | 11 (23.4%) | .890 |
| Alcohol drinking (%) | 38 (28.8) | 27 (31.8) | 11 (23.4) | .310 |
| Multiple aneurysms (%) | 7 (5.3%) | 3 (3.5%) | 4 (8.5%) | .414 |
| Complaints of symptoms | ||||
| Asymptomatic | 62 (47.0%) | 36 (42.2%) | 26 (55.3%) | .865 |
| Headache or dizziness | 44 (33.3%) | 33 (38.8%) | 11 (23.4%) | .085 |
| Cranial nerve palsy | 26 (19.7%) | 16 (18.8%) | 10 (21.3%) | .820 |
| Hypertension (%) | 57 (43.2%) | 44 (51.8%) | 13 (27.7%) | .013 |
| Diabetes (%) | 13 (9.8%) | 10 (11.8%) | 3 (6.4%) | .320 |
| Hyperlipidemia (%) | 12 (9.1%) | 8 (9.4%) | 4 (8.5%) | 1 |
| Cerebral infarction (%) | 17 (12.9%) | 13 (15.3%) | 4 (8.5%) | .265 |
| Coronary heart disease (%) | 9 (6.8%) | 5 (5.9%) | 4 (8.5%) | .720 |
aUnpaired t-test, chi-square test, or Fisher’s exact test as appropriate.
Table 2.
Aneurysm and treatment-related characteristics.
| Variable | Aneurysms with braided stents | p-Value a | ||
|---|---|---|---|---|
| Total (n = 139) | LVIS (n = 88) | Pipeline (n = 51) | ||
| Size, mm | 12.8 ± 9.0 | 9.8 ± 5.9 | 18.1 ± 11.0 | <.001 |
| Ruptured (%) | 41 (29.4%) | 30 (34.1%) | 11 (21.6%) | .119 |
| Non-saccular morphology (%) | 74 (53.2%) | 46 (52.3%) | 28 (54.9%) | .860 |
| Location | .002 | |||
| Basilar apex | 30 (21.6%) | 26 (29.5%) | 4 (7.8%) | |
| Basilar trunk | 30 (21.6%) | 13 (14.8%) | 17 (33.3) | |
| Vertebrobasilar junction | 79 (56.8%) | 49 (56.7%) | 30 (58.9%) | |
| Pre-treatment mRS | ||||
| 0–2 | 132 (95.0%) | 82 (93.2%) | 50 (98.0%) | .423 |
| 3–5 | 7 (5.0%) | 6 (6.8%) | 1 (2.0%) | |
| Overlapping stents (%) | 21 (15.9%) | 13 (15.3%) | 8 (17.0%) | .950 |
| Adjunctive coiling (%) | 108 (77.7%) | 83 (94.3%) | 25 (49%) | <.001 |
| Immediate complete/near-complete occlusion (%) | 82 (59%) | 72 (81.8%) | 10 (19.6%) | <.001 |
| Procedure time, min b | 90 (50) | 105 (40) | 85 (35) | .012 |
| Post-treatment mRS (%) | .405 | |||
| 0–2 | 119 (85.6%) | 77 (87.5%) | 42 (82.4%) | |
| 3–5 | 20 (14.4%) | 11 (12.5%) | 9 (17.6%) | |
| Follow-up time, m c | 7.6 ± 4.7 | 8.4±5.2 | 5.9±2.9 | .019 |
| Follow-up complete/near-complete occlusion (%) | 96 (69.1%) | 62 (70.5%) | 34 (66.7%) | .641 |
| Recurrence (%) | 15 (10.8) | 15 (17.0%) | 0 | .002 |
| Parent artery stenosis (%) | 3 (2.2%) | 0 | 3 (5.9%) | .09 |
| Branch artery occlusion (%) | 11 (7.9%) | 6 (6.8%) | 8 (15.7%) | .142 |
| Follow-up mRS (%) | .790 | |||
| 0–2 | 116 (83.5%) | 74 (84.1%) | 42 (82.4%) | |
| 3–5 | 23 (16.5%) | 14 (15.9%) | 9 (17.6%) | |
| Ischemic complication (%) d | 36 (25.9%) | 26 (29.5%) | 10 (19.6%) | .197 |
| Periprocedural period | 30 (21.6%) | 22 (25%) | 8 (15.7%) | .198 |
| Follow-up period | 14 (10.1%) | 11 (12.5%) | 3 (5.9%) | .212 |
| Hemorrhage complication (%) | 1 (0.7%) | 1 (1.1%) | 0 | 1 |
| Mass effect symptoms (%) | 13 (9.4%) | 4 (4.5%) | 9 (17.6%) | .024 |
| Death (%) | 7 (5.0%) | 3 (3.4%) | 4 (7.8%) | .261 |
aUnpaired t-test, Mann–Whitney U test, chi-square test, or Fisher’s exact test as appropriate.
bData are expressed as median (quartile).
cExcluding four patients who died during the post-operation hospitalization.
dEight patients with Pipeline and seven patients with LVIS suffered from ischemic complications in both the periprocedural period and the follow-up.
The aneurysm occlusion status was classified as complete occlusion (100%), near-complete occlusion (90%–99%), or incomplete occlusion (<90%). The aneurysm occlusion status before the operation, immediately after the operation, and during follow-up, parent artery stenosis (≥50% obstruction of the maximum vessel caliber), and branch artery occlusion at follow-up were documented. Postoperative CT or MRI examination was performed when patients develop hemiparesis, headache, blurred vision, and/or other neurological symptoms after the operation. Complications were defined as any procedure-related bleeding, infarction, or mass effect leading to neurologic deficit or death. Each complication was stratified as periprocedural (within ≤14 days) or during follow-up (>14 days). Ischemic complications were defined as any symptomatic in-stent thrombosis, thromboembolic events, transient ischemic attack, and cerebral infarcts associated with stent treatment. The modified Rankin scale (mRS) was used to assess the functional outcomes. All results were evaluated by two experienced neurointerventionists without interference.
Antiplatelet management and technical details
Before the procedure, the antiplatelet regimen included aspirin (100 mg daily) and clopidogrel (75 mg daily) for at least 5 days. Platelet function testing was performed by turbidimetric light transmittance aggregometry (LTA) at admission, and some clopidogrel non-responders with the maximal platelet aggregation response >50% to adenosine diphosphate (ADP) and >20% to arachidonic acid (AA)were transitioned to ticagrelor (90 mg twice daily) while continuing aspirin (100 mg daily). For patients with acutely ruptured aneurysms, a dual antiplatelet loading dose (300 mg aspirin and 300 mg clopidogrel once) was used before the procedure. The choice of treatment modality (LVIS or Pipeline) depended on the subjective judgment of interventionalists. All aneurysms were treated with patients under general anesthesia, via the transfemoral approach and with systemic heparin administration. Using the native and road map images, we deployed the stent to reconstruct the parent artery. The Pipeline stent was delivered through a Marksman or Phenom 27 microcatheter (Medtronic, Minneapolis, Minnesota, USA), while the LVIS stent was deployed through a Headway-21 (MicroVention, Aliso Viejo, CA, USA) microcatheter in all cases. The “jailing” catheter technique was performed for adjunctive coil placement. LVIS stent-assisted coiling was considered if the aneurismal sac contained sufficient space for coils. FD plus coiling was considered in the following two situations: (1) unstable aneurysm prone to rupture (i.e., with irregular shape or daughter blebs) and (2) max-size≥10 mm to promote earlier occlusion and avoid the low but catastrophic risk of delayed aneurysmal hemorrhage. Overlapping stents were considered if a single stent was insufficient to cover the entire lesion segment of dissecting and fusiform aneurysms. After the procedure, patients with LVIS received dual antiplatelet therapy for 3 months and then continued aspirin for at least 3 months. Patients with Pipeline received dual antiplatelet therapy for 6 months and then continued aspirin for at least 6 months.
Statistical analyses
Statistical analysis was performed with the R studio statistical software package. Nominal variables, including rates of aneurysm occlusion and complications, were presented as numbers and percentages of cases. Categorical variables were compared using the chi-square test or Fisher’s exact test, and continuous variables were compared using the unpaired t test or Mann–Whitney U test. p < .05 was considered statistically significant. PSM using the nearest neighbor method was performed, controlling for age, sex, hypertension, aneurysm size, shape (saccular/non-saccular), location (basilar apex/basilar trunk/VBJ), and duration of follow-up. Based on a caliper defined as 0.2SD of the logit of the PS, a 1–1 matching without replacement was executed.
Results
Baseline patient and aneurysm characteristics
Baseline patient characteristics are presented in Table 1, and aneurysm characteristics, treatment factors, and clinical outcomes are displayed in Table 2. With a mean age of 50.3 years, 62.1% were female. A total of 62 (47.0%) were asymptomatic patients, 44 (33.3%) had headache and dizziness, and 26 (19.7%) had central nerve palsy. 70.6% of aneurysms were unruptured. The aneurysm shape was saccular (46.8%) or non-saccular (53.2%). The mean aneurysm size was 12.8 ± 9.0 mm, and 15.8% of the aneurysms were ≥20 mm. The vertebrobasilar junction (56.8%) was the most common location followed by the basilar apex (21.6%) and the basilar trunk (21.6%). The Pipeline was used in 51 (36.7%) procedures and 25 (49.0%) with plus coiling. LVIS was used in 88 (63.3%) procedures, and 83 (94.3%) had additional dense-packed coils.
Four patients died during the postoperation hospitalization, and the remaining 128 patients completed follow-up (mean 7.6 ± 4.7 months). Angiographic complete or near-complete occlusion was reported in 69.1% of aneurysms. Ischemic complications occurred in 36 (25.9%) patients and usually (30/36, 83.3%) occurred during the periprocedural period. Only one patient (1.1%) experienced a hemorrhage (brain stem hematoma) during the follow-up period. Mass effect symptoms occurred in 13 (9.4%) patients. Patients’ complications were often relieved after drug treatments, and good functional outcomes (mRS 0–2) were achieved in 116 (83.5%) patients. The poor functional outcome rate (mRS 3–5) and mortality were 16.5% and 5%, respectively.
A total of 55 (39.6%) patients had platelet function testing before treatment, 15 of them were switched from clopidogrel to ticagrelor therapy, and 13 (23.6%) patients suffered postoperative ischemic complications. Among the 84 patients without platelet function testing, 23 (27.4%) had postoperative ischemic complications. There was no significant difference in the occurrence of postoperative ischemic complications between the platelet function testing group and the non-platelet function testing group (p = .662).
Unmatched comparison of LVIS and Pipeline
Compared with the LVIS group, the Pipeline group was significantly younger and had fewer patients with hypertension (Table 1), with larger size for aneurysms (9.8 ± 5.9 mm vs 18.1 ± 11.0 mm, p < .001); had more stent-only procedures and lower immediate angiographic occlusion rate (5.7% vs 51%, p < .001; 81.8% vs 19.6%, p < .001); was associated with a shorter procedure time (median 85 (35) versus 105 (40), p = .012);and had lower recurrence rate but more postoperative mass effect (17.0% vs 0, p = .002; 4.5% vs 17.6%, p = .024). LVIS was more often used to treat basilar apex aneurysms. Other factors showed no significant differences (p > .05, Table 2).
Comparison of LVIS and Pipeline after PSM
Controlling for age, sex, hypertension, aneurysm size, shape (saccular/non-saccular), location (basilar apex/basilar trunk/VBJ), and duration of follow-up, matching was successful for 30 LVIS and Pipeline aneurysm pairs. After PSM, Table 3 showed the comparison of the two groups and some results remained significant: (1) An overwhelming majority of 96.7% cases had adjunctive coiling in the LVIS group, but doctors preferred to use Pipeline alone (63.6%); (2) compared with the LVIS group, the Pipeline group had lower immediate occlusion rate (90% vs 26.7%; p < .001) but similar follow-up occlusion rates (76.7% vs 73.3%; p = .766); (3) favorable functional outcomes (mRS 0–2) achieved in most patients (LVIS 86.7% vs Pipeline 90.0%; p = 1). Meanwhile, some results changed: (1) Compared with the LVIS group, the Pipeline group had a lower incidence of ischemic complications; however, this finding was very close to but did not reach statistical significance (30% vs 10%; p = .053). (2) There was no statistically significant difference in the comparison of procedure times (mean 95.2 ± 36.8 min vs 90 ± 24.5 min, p = .524), recurrence rates (13.3% vs 0, p = .112), and mass effect rates (13.3% vs 3.3%, p = .353) between the two groups.
Table 3.
Clinical and angiographic characteristics after 1:1 matching by PSM.
| Variable | Aneurysms with braided stents | p-Value a | |
|---|---|---|---|
| LVIS (n = 30) | Pipeline (n = 30) | ||
| Size, mm | 12.8 ± 7.3 | 13.3 ± 6.9 | .78 |
| Pre-treatment mRS | .492 | ||
| 0–2 | 28 (93.3%) | 30 (100%) | |
| 3–5 | 2 (6.7%) | 0 | |
| Overlapping stents (%) | 4 (13.3%) | 1 (3.3%) | .339 |
| Adjunctive coiling (%) | 29 (96.7%) | 11 (36.7%) | <.001 |
| Immediate complete/near-complete occlusion (%) | 27 (90%) | 8 (26.7%) | <.001 |
| Procedure time, min | 95.2±36.8 | 90±24.5 | .524 |
| Post-treatment mRS (%) | 1 | ||
| 0–2 | 27 (90%) | 28 (93.3%) | |
| 3–5 | 3 (10%) | 2 (6.7%) | |
| Follow-up complete/near-complete occlusion (%) | 23 (76.7%) | 22 (73.3%) | .766 |
| Recurrence (%) | 4 (13.3%) | 0 | .112 |
| Follow-up mRS (%) | |||
| 0–2 | 26 (86.7%) | 27 (90%) | 1 |
| 3–5 | 4 (13.3%) | 3 (10%) | |
| Parent artery stenosis (%) | 0 | 1 (3.3%) | 1 |
| Branch artery occlusion (%) | 1 (3.3%) | 1 (3.3%) | 1 |
| Ischemic complication (%) | 9 (30.0%) | 3 (10.0%) | .053 |
| Hemorrhage complication (%) | 1 (3.3%) | 0 | 1 |
| Mass effect symptoms (%) | 4 (13.3%) | 1 (3.3%) | .353 |
| Death (%) | 1 (3.3%) | 0 | 1 |
aUnpaired t-test, chi-square test, or Fisher’s exact test as appropriate.
Comparison of LVIS and Pipeline at different locations
The location of BA aneurysms sometimes dictates the technique choice. So we compared the two treatment strategies at three different locations (basilar apex/basilar trunk/VBJ, Table 4). For treatment factors, all basilar apex aneurysms were treated with adjunctive coiling in the LVIS and Pipeline groups. For basilar apex and basilar trunk aneurysms, the two stents showed similar rates of complete or near-complete occlusion. For VBJ aneurysms, the rate of complete or near-complete occlusion was higher in the Pipeline group than in the LVIS group (86.7% vs 59.2%; p = .012) (Figures 1–4). The characteristics related to VBJ aneurysm occlusion were then further analyzed, but no statistical significance was found except for the use of Pipeline (Table 5). Due to the lack of sufficient cases, other outcome factors, including complications and recurrence, could not be compared.
Figure 2.
A non-saccular aneurysm received LVIS treatment. (a) The vertebral angiography showed a vertebrobasilar dolichoectasia located in the left vertebrobasilar junction, and the length was 23.2 mm. (b) The microcatheter was displaced from the left vertebral artery to the basilar artery. (c) An LVIS stent (5.5 × 25 mm) was deployed in the left vertebral artery and basilar artery. (d) The vertebrobasilar dolichoectasia still existed at 12-month follow-up.
Figure 3.
A saccular aneurysm received Pipeline flow diverter treatment. (a) Three-D vertebral angiography construction showed an aneurysm located at the vertebrobasilar junction, and the maximum size was 15.8 mm. (b) A Pipeline flow diverter (4.25 × 25 mm) was displaced from the basilar artery to the right vertebral artery and three coils were loosely packed. (c) The aneurysm achieved complete occlusion at 12-month follow-up. (d) The unsilhouette picture showed clearly that the Pipeline flow diverter adhered the parent artery and the aneurysm lumen was filled with coils.
Table 4.
Aneurysm occlusion at different locations.
| Location | Aneurysms with complete or near-complete occlusion | p-Value a | ||
|---|---|---|---|---|
| Total | LVIS group | Pipeline group | ||
| Basilar apex (n = 30) | 25 (83.3%) | 23 (88.5%) | 2 (50%) | .119 |
| Basilar trunk (n = 30) | 19 (63.3%) | 10 (76.9%) | 9 (52.9%) | .259 |
| Vertebrobasilar junction (n = 79) | 55 (69.6%) | 29 (59.2%) | 26 (86.7%) a | .012 |
aChi-square test or Fisher’s exact test as appropriate.
Figure 1.
A saccular aneurysm received LVIS treatment. (a) The vertebral angiography showed an aneurysm located at the left vertebrobasilar junction, and the maximum size was 6.5 mm; the slim vertebral artery may be an anatomic variation. (b) Immediate posttreatment angiography showed complete occlusion of the aneurysm. (c) An LVIS stent (3.5 × 20 mm) was deployed and four coils were densely packed. (d) The follow-up angiography showed only near-complete occlusion of aneurysm at 6 months.
Figure 4.
A non-saccular aneurysm received Pipeline flow diverter treatment. (a) The 3-D vertebral angiography construction showed a basilar dissecting aneurysm located at the vertebrobasilar junction, and the maximum size was 14.3 mm. (b) A Pipeline flow diverter (4.25 × 35 mm) was implanted from the basilar artery to the right vertebral artery. (c)–(d) The anteroposterior and lateral views of computerized tomography angiography (CTA) showed that the aneurysm was completely occluded at 12-month follow-up.
Table 5.
Characteristics related to aneurysm occlusion at vertebrobasilar junction.
| Variable | Vertebrobasilar junction aneurysms | p-Value a | |
|---|---|---|---|
| Complete or near-complete occlusion (n = 52) | Incomplete occlusion (n = 27) | ||
| Age, y | 48.4 ± 16.2 | 51.0 ± 79 | .337 |
| Male (%) | 32 (61.5%) | 15 (55.6%) | .604 |
| Ruptured (%) | 20 (38.5%) | 9 (33.3%) | .654 |
| Size, mm | 13.4±9.3 | 15.0±7.9 | .447 |
| Non-saccular shape (%) | 35 (67.3%) | 21 (77.8%) | .331 |
| Adjunctive coiling (%) | 41 (78.8%) | 20 (74.1%) | .631 |
| Stent type | |||
| LVIS (%) | 29 (59.2%) | 20 (40.8%) | .012 |
| Pipeline (%) | 26 (86.7%) | 4 (13.3%) | |
| Follow-up time, m | 6.9 ± 3.6 | 8.1 ± 3.1 | .134 |
aUnpaired t-test or chi-square test as appropriate.
Discussion
Posterior circulation aneurysms have a higher risk of rupturing and becoming symptomatic.1,7 BA aneurysms constitute 50% to 80% of posterior circulation aneurysms. 3 Due to the increased number of perforating arteries and their supply to many vital structures, neuroendovascular surgeons have been cautious about their treatment strategy. Microsurgical treatment of these aneurysms is difficult and associated with higher morbidity and mortality rates. 6 Endovascular treatment has become the predominant treatment modality if it is technically feasible. With frequently large sizes and non-saccular shapes, stent-assisted coiling was usually the preferred choice before the advent of FDs. Additionally, the current prevalent use of FDs in the posterior circulation remains controversial. Experience with FDs for the treatment of BA aneurysms is limited to small clinical series, providing somewhat mixed results.2,5,8
This study reported the results of our 7-year experience in treating aneurysms at this rare and challenging location, including 139 consecutive cases. We focused our attention on the comparison of two types of braided stents (Pipeline FD vs low-profile LVIS stent). The novel findings are that, after PSM to achieve baseline covariate balance, both braided stents provided favorable occlusion rates and functional outcomes for the treatment of BA aneurysms; Pipeline FD treatment has achieved a significantly higher complete or near-complete occlusion rate of aneurysms (86.7%) than LVIS stent treatment at the site of VBJ, where aneurysms are often considered extremely challenging for endovascular treatment.9–11
At the BA location, as previously mentioned, the controversy over the application of FDs rests on the concern about covered perforators and thromboembolic complications. At the same time, there are many fusiform and dissecting aneurysms (53.2% in this study), which represent a unique challenge due to their high bleed risk and their morphology not being amenable to standard endovascular approaches.1,11,12 These lesions usually require vascular remodeling and a strategy of stent-assisted coiling or FDs. According to this, with a 17%–23% metal coverage to allow for more flow into the perforators and a certain degree of flow-diverting effect in theory, LVIS stent with adjunctive coiling was applied as a therapeutic strategy. With a 30%–35% metal coverage and a strong flow-diverting effect, Pipeline FDs alone or plus coiling were tried too. The results proved that LVIS and Pipeline groups achieved similar complete or near-complete occlusion rates (76.7% vs 73.3%) and favorable functional outcomes (86.7% vs 90.0%) in treating BA aneurysms, respectively. These results are similar to the published reports. Lozier et al. 13 reviewed a series of 495 posterior circulation aneurysms treated with coil embolization in 12 institutions; 82% of aneurysms arose near the basilar apex; of coiled basilar apex aneurysms, 72.9% achieved complete or near-complete aneurysm occlusion and 84.0% achieved functional independence at follow-up. 13 Henkes et al. 14 analyzed 316 BA bifurcation aneurysms treated using endovascular coil occlusion, 70% achieved a 90 to 100% occlusion rate at 19-month follow-up and 81% achieved independent functional outcomes. 14 Adeed et al. 5 reviewed 378 procedures in treating posterior circulation aneurysms with Pipeline or the FRED FD. At a mean follow-up of 13 months, complete or near-complete occlusion was achieved in 84.5% aneurysms, and mRS 0–2 was achieved in 86.8% aneurysms. 2
In this study, the ischemic complication rate was reduced in the Pipeline group but did not reach statistical significance (30% vs 10%; p = .053), and there were non-significant decreased mass effect symptoms (13.3% vs 3.3%; p = .353) in the Pipeline group. Ischemic complications usually (83.3%) occurred during the periprocedural period. There was no difference in the rate of parent artery stenosis and branch occlusion between the two stents in this study. This may be because the available sample size is still not large enough, resulting in an incomplete analysis of risk factors for complications. A larger sample might enhance the study to detect significant differences between the groups.
VBJ aneurysms have poor natural history and pose significant treatment challenges regardless of treatment techniques.9–11 Our study included 79 VBJ aneurysms, of which a particularly high proportion of non-saccular shape (56 cases; 70.9%) and a male preponderance were noted. As mentioned above, the treatment of non-saccular aneurysms requires the reconstruction of the diseased vessel. FDs theoretically can offer a particular advantage; however, the early experience was notable for several reports of poor outcomes in posterior circulation aneurysms.8,15 With expanded experience, neurosurgeons tried to avoid overlapping FDs if possible and tended to add additional coils to treat posterior circulation aneurysms, and then better to excellent results were reported.5,12,15 Graziano et al. 10 conducted a meta-analysis of endovascular treatments for VBJ aneurysms and found that 13/17 (76.5%) VBJ aneurysms with FDs achieved a good outcome. Bhogal et al. 12 reported 56 patients with 58 non-saccular posterior circulation aneurysms treated with Pipeline or p64 FD, and the majority of the lesions were VBJ in location (44.8%). The mortality rate was only 4% (1 death), and 38/46 (82.6%) aneurysms achieved a complete or near-complete occlusion rate with a mean of 25.2 months follow-up. 12 Griessenauer et al. 16 studied 149 consecutive posterior circulation aneurysms treated with Pipeline in 11 centers, of which 64.4% were non-saccular. At a median follow-up of 12 months, 97.8% of the dissecting/blister aneurysms and 91.9% of the fusiform aneurysms achieved complete or near-complete occlusion. Pipeline FD is preferentially used for the treatment of fusiform, dissecting, and blister aneurysm morphologies. 16 In this study, a particularly favorable outcome of Pipeline FD treatment was found at the VBJ location, and the complete or near-complete aneurysm occlusion rate was much higher than LVIS stent (86.7% vs 59.2%; p = .012). Our results are consistent with these prior studies, indicating that the use of FDs for the treatment of complex aneurysms at the VBJ site can achieve excellent embolization results. The theoretical degree of flow-diverting properties of low-profile braided LVIS stent did not achieve a good aneurysm occlusion rate for this special subtype.
Limitations
First, the retrospective observational cohort study design and relatively small sample size may lead to some degree of bias. Secondly, only 39.6% of patients in this study underwent platelet function tests to adjust their dual-antiplatelet therapy before stent treatments. Although there was no significant difference in the occurrence of postoperative ischemic complications between the platelet function testing group and the non-platelet function testing group, the relationship between this test and hemorrhagic complications cannot be analyzed because of insufficient sample size. Finally, short follow-up time may affect the follow-up results and the incidence of complications, and the complete embolism rate in the Pipeline cohort is likely to increase with longer follow-up time.
Conclusion
This study compared two braided stents for the treatment of BA aneurysms and used data from 139 consecutive aneurysms at this specific location in two medical centers. After propensity score matching, both braided stents provided a favorable occlusion rate and functional outcome for the treatment of BA aneurysms. The aneurysm at the VBJ location has a high proportion of non-saccular morphology and often requires disease vascular reconstruction. Pipeline FDs achieved a relatively higher aneurysm occlusion rate at this location, while LVIS stents did not perform well in this regard with a lower aneurysm occlusion rate.
Footnotes
Author contributions: Methodology: Bin Luo, Chao Wang, and Ying Zhang; formal analysis and investigation: Bin Luo, Chao Wang, and Ying Zhang; writing—original draft preparation: Bin Luo and Chao Wang; writing—review and editing: Jian Liu, Yisen Zhang, Kun Wang, Wenqiang Li, and Ying Zhang; funding acquisition: Jian Liu and Ying Zhang; supervision and approved the version: Ying Zhang.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The National Natural Science Foundation of China (grant numbers: 82072036, 82272092, and 82372058); Capital’s Funds for Health Improvement and Research (grant numbers: 2022-1-2041); and Summit Talent Program grant number: (DFL20220504).
Ethical statement
This study was conducted according to the World Medical Association Declaration of Helsinki. The study protocol was reviewed and approved by the Ethics Committee of Beijing Tiantan Hospital, and the approval number given by the ethical board was KY 2018-098-02. Individual patients’ written informed consent was waived owing to the retrospective nature of the study.
ORCID iD
Ying Zhang https://orcid.org/0000-0002-5812-372X
References
- 1.Dawkins D, Sayyahmelli S, Baskaya MK. Posterior circulation aneurysms. In: Kato Y, Ansari A. (eds) Cerebrovascular surgery: controversies, standards and advances. Cham: Springer International Publishing, 2022, pp. 55–77. [DOI] [PubMed] [Google Scholar]
- 2.Marlin ES, Ikeda DS, Shaw A, et al. Endovascular treatment of basilar aneurysms. Neurosurg Clin 2014; 25: 485–495. 2014/07/06 DOI: 10.1016/j.nec.2014.04.007. [DOI] [PubMed] [Google Scholar]
- 3.Rangel-Castilla L, Munich SA, Levy EI. Craniotomy is over for basilar artery aneurysms. In: Veznedaroglu E. (ed) Controversies in vascular neurosurgery. Cham: Springer International Publishing, 2016, pp. 93–108. [Google Scholar]
- 4.Iosif C, Biondi A. Braided stents and their impact in intracranial aneurysm treatment for distal locations: from flow diverters to low profile stents. Expet Rev Med Dev 2019; 16: 237–251. DOI: 10.1080/17434440.2019.1575725. [DOI] [PubMed] [Google Scholar]
- 5.Adeeb N, Dibas M, Griessenauer CJ, et al. Learning curve for flow diversion of posterior circulation aneurysms: a long-term international multicenter cohort study. AJNR Am J Neuroradiol 2022; 43: 1615–1620. DOI: 10.3174/ajnr.A7679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Cagnazzo F, Cappucci M, Lefevre PH, et al. Treatment of intracranial aneurysms with self-expandable braided stents: a systematic review and meta-analysis. AJNR Am J Neuroradiol 2018; 39: 2064–2069. DOI: 10.3174/ajnr.A5804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Adeeb N, Ogilvy CS, Griessenauer CJ, et al. Expanding the indications for flow diversion: treatment of posterior circulation aneurysms. Neurosurgery 2020; 86: S76–S84. DOI: 10.1093/neuros/nyz344. [DOI] [PubMed] [Google Scholar]
- 8.Siddiqui AH, Abla AA, Kan P, et al. Panacea or problem: flow diverters in the treatment of symptomatic large or giant fusiform vertebrobasilar aneurysms. J Neurosurg 2012; 116: 1258–1266. DOI: 10.3171/2012.2.JNS111942. [DOI] [PubMed] [Google Scholar]
- 9.Alwakeal A, Shlobin NA, Golnari P, et al. Flow diversion of posterior circulation aneurysms: systematic review of disaggregated individual patient data. AJNR Am J Neuroradiol 2021; 42: 1827–1833. DOI: 10.3174/ajnr.A7220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Graziano F, Ganau M, Iacopino DG, et al. Vertebro-basilar junction aneurysms: a single centre experience and meta-analysis of endovascular treatments. NeuroRadiol J 2014; 27: 732–741. DOI: 10.15274/NRJ-2014-10100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kiyofuji S, Graffeo CS, Perry A, et al. Meta-analysis of treatment outcomes of posterior circulation non-saccular aneurysms by flow diverters. J Neurointerventional Surg 2018; 10: 493–499. DOI: 10.1136/neurintsurg-2017-013312. [DOI] [PubMed] [Google Scholar]
- 12.Bhogal P, Pérez MA, Ganslandt O, et al. Treatment of posterior circulation non-saccular aneurysms with flow diverters: a single-center experience and review of 56 patients. J Neurointerventional Surg 2017; 9: 471–481. DOI: 10.1136/neurintsurg-2016-012781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lozier AP, Connolly ES, Jr., Lavine SD, et al. Guglielmi detachable coil embolization of posterior circulation aneurysms: a systematic review of the literature. Stroke 2002; 33: 2509–2518. DOI: 10.1161/01.str.0000031928.71695.a9. [DOI] [PubMed] [Google Scholar]
- 14.Henkes H, Fischer S, Mariushi W, et al. Angiographic and clinical results in 316 coil-treated basilar artery bifurcation aneurysms. J Neurosurg 2005; 103: 990–999. DOI: 10.3171/jns.2005.103.6.0990. [DOI] [PubMed] [Google Scholar]
- 15.Siddiqui AH, Kan P, Abla AA, et al. Complications after treatment with pipeline embolization for giant distal intracranial aneurysms with or without coil embolization. Neurosurgery 2012; 71: E509–E513. DOI: 10.1227/NEU.0b013e318258e1f8. [DOI] [PubMed] [Google Scholar]
- 16.Griessenauer CJ, Enriquez-Marulanda A, Taussky P, et al. Experience with the pipeline embolization device for posterior circulations aneurysms: a multicenter cohort study. Neurosurgery 2020; 87: 1252–1261. DOI: 10.1093/neuros/nyaa277. [DOI] [PubMed] [Google Scholar]