Pipeline embolization of proximal middle cerebral artery aneurysms: A multicenter cohort study

Abstract

Background and purpose

Flow diversion of aneurysms located in the M1 segment and middle cerebral artery bifurcation with Pipeline embolization device is sometimes performed, but further study is needed to support its regular use in aneurysm treatment. Here, we report measures of safety and efficacy for Pipeline embolization in the proximal middle cerebral artery in a multi-center cohort.

Materials and methods

Clinical and angiographic data of eligible patients were retrospectively obtained from participating centers and assessed for key clinical and angiographic outcomes. Additional details were extracted for patients with complications.

Results

In our multi-center cohort, complete aneurysm occlusion was achieved in 71% (17/24) of treated aneurysms. There were no deaths or disabling strokes, but non-disabling ischemic strokes occurred in 8% (2/24) of patients. For aneurysms in the M1 segment, complete aneurysm occlusion was observed in 75% (12/16) of aneurysms, aneurysm volume reduction was observed in 100% (16/16) of aneurysms, and non-disabling ischemic strokes occurred in 13% (2/16) of patients. For aneurysms at the middle cerebral artery bifurcation, complete aneurysm occlusion was observed in 63% (5/8) of aneurysms, aneurysm volume reduction occurred in 88% (7/8) of aneurysms, and ischemic or hemorrhagic complications occurred in 0% (0/8) of patients.

Conclusion

Pipeline embolization of cerebral aneurysms in the M1 segment and middle cerebral artery bifurcation demonstrated a 71% rate of complete aneurysm occlusion. There were no deaths or disabling strokes, but there was an 8% rate of non-disabling ischemic strokes.

Introduction

Flow diversion is a common treatment method for select brain aneurysms.^1–3 The Pipeline Embolization Device (PED) is the most frequently used flow diverting stent and demonstrates high levels of safety and efficacy in approved locations. ² , ⁴ , ⁵ Off-label performance of the PED in other regions is varied, with studies suggesting favorable safety and efficacy in anterior cerebral arteries, unfavorable efficacy in fetal origin posterior communicating arteries, and unfavorable safety in the posterior circulation.^6–9 Certain aneurysms may prompt consideration of off-label PED use, including middle cerebral artery (MCA) aneurysms involving the M1 segment or the MCA bifurcation. Aneurysms of the MCA have traditionally been treated via open microsurgical clipping, but endovascular approaches are sometimes utilized. ¹⁰ PED use in the proximal MCA may be challenging given the regional anatomy and arterial branches, including small lenticulostriate arteries, that may present similar challenges as flow diversion in the posterior circulation.^11–14

While flow diversion in the proximal MCA has been described previously, further study is necessary.^15–20 Some of these studies have reported high efficacy in the M1 segment, ¹⁹ , ²¹ but reports on safety in this region vary. ¹⁴ , ²² This ambiguity highlights the need for the acquisition of targeted safety and efficacy data for PED use in the proximal MCA. In this work, we review safety- and efficacy-related outcomes of flow diversion with PED in the M1 and MCA bifurcation in a multi-center patient cohort.

Methods

Patient selection

Each participating center received separate institutional review board for this study. Consecutive patients undergoing proximal MCA aneurysm treatment with PED at two high-volume neurovascular centers from October 2011 to March 2020 were reviewed. Patients who had undergone previous aneurysm treatment were included in this study.

Embolization procedure

Informed consent for each procedure was obtained for all patients per clinical routine before treatment. Prior to treatment, patients were provided with dual antiplatelet therapy that typically comprised aspirin plus clopidogrel or ticagrelor with dose titration based on VerifyNow assay. ²³ Intraprocedural systemic heparinization was provided to attain elevation of activated clotting time of approximately 2.5 times above patient baseline. All patients underwent embolization via a transfemoral or transradial route in the supine position in a biplane neurointerventional suite. Post-procedural care, including the use of closure devices and access site monitoring, was performed per clinical routine. Initial angiographic follow-up was typically performed approximately 6 months post-treatment. Dual antiplatelet therapy was generally continued for 6–12 months, followed by indefinite aspirin monotherapy.

Data acquisition and classification

Patient characteristics (age, sex, major comorbidities, indication for treatment), aneurysm characteristics (location, size, shape, presentation, parent vessel size), peri-procedural details (antiplatelet dosage and activity corresponding to the available PRU closest in time to the treatment angiogram, patient level of function), treatment details (retreatment status, PED delivery system, device size, challenges in deployment, immediate procedural complications), angiographic follow-up (aneurysm occlusion, in-stent stenosis, patency of covered side branches), clinical follow-up (patient level of function, ischemic or hemorrhagic complications), and findings from available cross-sectional imaging were collected at both centers. Assessed comorbidities included hypertension, diabetes mellitus, hyperlipidemia, and recent smoking history. Patients who quit smoking more than 6 months before PED placement did not have smoking counted as a comorbidity. Total duration of follow-up was determined as the time from PED placement to the most recent angiography. Any angiography performed 6 ± 2 months after PED placement is considered 6-month follow-up angiography. No patient data was omitted from the study, and patients not presenting for follow-up were still included in this study to acquire relevant periprocedural data. Aneurysm size is reported based on its largest dimension recorded prior to treatment, and aneurysm volume reduction and occlusion was determined using the O’Kelly-Marotta grading scale measuring degree of angiographic contrast filling, where total filling is defined by >95% filling of the aneurysm, subtotal filling is defined by 5–95% filling of the aneurysm, entry remnant is defined by <5% filling of the aneurysm, and occlusion is defined by 0% filling of the aneurysm. ²⁴ Major side branches that could be consistently identified on patient angiograms were included in evaluation of side branch patency. In-stent stenosis is reported as a percentage decrease in luminal caliber relative to normal distal artery, or as “occluded” if angiography demonstrated no flow in the PED. Aneurysm occlusion status was reported for both 6-month follow-up angiography and longest follow-up in patients presenting for later imaging. Rates of aneurysm occlusion and ischemic or hemorrhagic complication were calculated on a per-aneurysm and per-patient basis, respectively. In-stent stenosis rate and technical complication rate were calculated on a per-treatment basis. Baseline and post-procedural level of function was measured with the modified Rankin Scale (mRS). ²⁵ All data were aggregated into dedicated PED databases at both participating centers, but only de-identified data were shared with the organizing site.

Results

Patients and aneurysms

In total, 25 patients with 25 proximal MCA aneurysms were treated in 27 procedures. Demographics of this cohort are outlined in Table 1. Nineteen aneurysms had saccular morphology and 6 were fusiform, with a mean aneurysm size of 9.4 mm (range 3.3–20 mm). Seventeen treated aneurysms were located in the M1 segment, and 8 were located at the MCA bifurcation. Ten aneurysms had been treated previously. PEDs used had diameters ranging from 2.5–5.0 mm and lengths ranging from 10–35 mm. Two aneurysms were initially treated with 2 PEDs deployed in the same procedure, while the remainder had one device placed per procedure. Two patients presented with symptoms, one due to aneurysm rupture and the other with seizures.

Table 1.

Demographics of our cohort. Rates reported as n (%).

Mean age	53.7 years (range 16–72 years)
Sex (female)	15 (60%)
Comorbidities
Hypertension	25 (72%)
Smoker	8 (32%)
Hyperlipidemia	8 (32%)
Diabetes	6 (24%)
None	3 (12%)
Aneurysm status
Unruptured, asymptomatica	23 (92%)
Unruptured, symptomatic	1 (4%)
Ruptured	1 (4%)
Previous treatment
Coiling	7 (28%)
Clipping	1 (4%)
PED	2 (8%)
None	15 (60%)

^aIncludes incidentally discovered aneurysms and asymptomatic recurrent aneurysms.

Clinical outcomes and complications

Technical complications occurred in 7% (2/27) of procedures, including 1 vasospasm induced by catheter navigation and 1 PED herniation into the aneurysm. Neither of these technical complications resulted in clinical sequelae.

Clinical follow-up was available in all but one patient treated with PED, for a total of 24 patients. Mean duration of clinical follow-up was 33 months (range 1–93 months). Ischemic or hemorrhagic complications were observed in 8% (2/24) of patients, both of which were ischemic infarcts following PED placement in an M1 segment. Neither of these complications resulted in permanent disability. The ischemic or hemorrhagic complication rate was 13% (2/16) for M1 aneurysms and 0% (0/8) for MCA bifurcation aneurysms. There were no deaths or disabling strokes following PED treatment.

Patient 4 experienced ischemic stroke secondary to PED thrombosis 4 months post-treatment. VerifyNow assay performed at time of stroke presentation returned results of 182 PRU and 645 ARU, suggesting adequate P2Y12 inhibition by clopidogrel. This patient was not treated with thrombolytic therapy and was discharged 2 days after presentation with a return to normal level of function (mRS 0). This patient’s aneurysm was occluded at follow-up imaging. No additional complications occurred.

Patient 11 experienced ischemic stroke with thrombosis of PED two weeks post-procedure. VerifyNow assay performed at time of stroke presentation returned results of 302 PRU and 420 ARU, suggesting inadequate P2Y12 inhibition by clopidogrel. IV and IA abciximab administration restored patency of the PED, and the patient was discharged 5 days after initial stroke presentation with a return to baseline level of function (mRS 0). This patient’s aneurysm was occluded at this time. No additional complications occurred.

Cross-sectional imaging follow-up was available in 18 patients, including Patient 4 and Patient 11. The remaining 16 patients were asymptomatic in the follow-up period, but 19% (3/16) had clinically silent infarcts identified with cross-sectional imaging.

Angiographic outcomes

Angiographic follow-up was available in all but one patient treated with PED, for a total of 24 patients. Mean duration of angiographic follow-up was 30 months (range 1–74 months). Follow-up data are shown in Table 2. Imaging from a representative patient is shown in Figure 1.

Table 2.

Aneurysm treatment and follow-up. Each line corresponds to a unique aneurysm.

Patient number	Aneurysm location	Aneurysm size (mm)	Aneurysm morphology	Parent vessel size (mm)	PED size (mm)	Pre-procedure PRU	mRS score (baseline, follow-up)	Complications	Clinical follow-up (months)	Angiographic follow-up (months)	6-month in-stent stenosis	6-month aneurysm patency	Final in-stent stenosis	Final aneurysm patency	Cross-sectional imaging (modality)
1	Right M1	10.0	Saccular	2.7	2.5 × 10	213	0,0	None	93	62	50%	Occluded	<10%	Occluded	N/A
2	Left M1	4.0	Fusiform	3.1	4.0 × 20	118	3,0	None	5	5	None	Occluded	None	Occluded	N/A
3	Right M1	6.0	Fusiform	2.5	2.75 × 16	262	0,0	Vasospasm	0	0	N/Aa	N/Aa	N/Aa	N/Aa	N/A
4	Left M1	11.9	Fusiform	3.2	4.5 × 16, 5.0 × 16b	150	0,0	PED thrombosis, ischemic stroke	27	27	None	Total filling	None	Occluded	Left basal ganglia infarct (CT)
5	Left M1	4.4	Saccular	2.4	3.0 × 12, 3.25 × 12c	162, 157	1,1; 1,1	None	71	71	None	Total filling	None	Subtotal filling	N/A
6	Right M1	4.2	Saccular	2.5	3.0 × 12	132	3,3	None	24	24	None	Occluded	None	Occluded	Silent right basal ganglia infarct (MRA, MRI)
7	Right M1	8.9	Saccular	2.7	2.5 × 10	121	3,3	None	24	24	None	Subtotal filling	None	Subtotal filling	No finding (MRA, MRI, CT)
8	Right M1	19.5	Saccular	2.6	3.0 × 14	71	1,2	None	60	15	None	Occluded	None	Occluded	N/A
9	Right M1	13.4	Saccular	3.0	3.25 × 14	95	0,0	None	46	32	60%	Total filling	None	Occluded	N/A
10	Left M1	20.0	Fusiform	3.3	4.0 × 35, 4.75 × 25b	231	2,2	None	7	7	None	Occluded	None	Occluded	N/A
11	Left M1	11.5	Saccular	2.8	3.25 × 14	122	0,0	PED thrombosis, ischemic stroke	1	1	N/Ae	N/Ae	Occluded	Occluded	Left basal ganglia infarct (CT)
12	Right M1	3.3	Saccular	2.4	2.5 × 10	182	2,2	None	7	7	None	Subtotal filling	None	Subtotal filling	No finding (CTA, CT)
13	Left M1	5.1	Saccular	2.4	2.5 × 10	126	0,0	None	27	27	None	Occluded	None	Occluded	No finding (CTA, MRA)
14	Right M1	5.1	Saccular	2.4	2.5 × 10	182	0,0	None	63	63	None	Subtotal filling	None	Occluded	No finding (CTA, MRA)
15	Right M1	20.0	Saccular	2.6	2.75 × 10	89	0,0	None	7	7	None	Occluded	None	Occluded	No finding (MRA, MRI)
16	Left M1	6.3	Fusiform	2.2	2.5 × 10	129	0,0	None	3	3	N/Ae	N/Ae	None	Subtotal filling	No finding (CTA, MRA, MRI)
17	Right M1	4.0	Saccular	2.2	2.5 × 12	118	1,1	None	60	60	None	Occluded	None	Occluded	Silent basal ganglia infarct (MRA)
18	Right MCA bifurcation	6.0	Fusiform	2.9	2.75 × 14, 3.0 × 18c	142, 164	3,3; 3,3	PED malpositionedd	67	67	None	Occluded	None	Occluded	No finding (MRA, MRI, CT)
19	Left MCA bifurcation	5.2	Saccular	2.8	2.75 × 14	104	0,1	None	74	74	None	Occluded	None	Occluded	No finding (CT)
20	Right MCA bifurcation	19.0	Saccular	2.4	2.5 × 14f	161	0,1	None	33	33	None	Entry remnant	None	Entry remnant	Silent cortical infarct (MRA)
21	Right MCA bifurcation	18.3	Saccular	3.1	3.25 × 20	Unknowng	0,0	None	6	6	None	Occluded	None	Occluded	No finding (MRA, MRI)
22	Right MCA bifurcation	6.0	Saccular	2.6	2.75 × 12	58	0,0	None	6	6	None	Occluded	None	Occluded	No finding (MRA)
23	Right MCA bifurcation	3.9	Saccular	2.5	2.75 × 14	195	2,1	None	24	24	None	Occluded	None	Occluded	No finding (MRA, MRI)
24	Right MCA bifurcation	10.1	Saccular	2.6	3.5 × 12	113	0,0	None	60	60	None	Total filling	None	Total filling	No finding (MRA, MRI)
25	Right MCA bifurcation	9.5	Saccular	2.6	3.0 × 14	Unknowng	0,0	None	6	6	None	Subtotal filling	None	Subtotal filling	No finding (MRA)

^aDid not complete any post-treatment follow up.

^bBoth PEDs placed in the same session.

^cPEDs placed in separate sessions.

^dFirst PED herniated into aneurysm, second PED was deployed through first PED into aneurysm outflow tract.

^eDid not complete 6-month follow up angiography.

^fPED herniated into aneurysm and was repositioned during treatment session.

^gSample collected and reported at another facility.

Figure 1.

Angiographic images in Patient 9. (a) Pre-treatment angiography and (b) pre-treatment 3 D reconstruction show a 13.4 × 12.0 mm aneurysm of the right M1 segment. (c) Immediate post-treatment angiography following deployment of 3.25 × 14 mm PED showing contrast stasis within the aneurysm. (d) 32-month follow-up angiography demonstrates complete occlusion of the aneurysm.

At 6-month angiographic follow-up, complete aneurysm occlusion was observed in 59% (13/22) of aneurysms, comprising 57% (8/14) of M1 aneurysms and 63% (5/8) of MCA bifurcation aneurysms. Aneurysm volume reduction was observed in 82% (18/22) of aneurysms, comprising 79% (11/14) of M1 aneurysms and 88% (7/8) of MCA bifurcation aneurysms. In-stent stenosis or occlusion was observed following 8% (2/24) of treatments at 6-month follow-up imaging.

At final angiographic follow up, complete aneurysm occlusion was observed in 71% (17/24) of aneurysms, comprising 75% (12/16) of M1 aneurysms and 63% (5/8) of MCA bifurcation aneurysms. Aneurysm volume reduction was observed in 96% (23/24) of aneurysms, comprising 100% (16/16) of M1 aneurysms and 88% (7/8) of MCA bifurcation aneurysms. In-stent stenosis or occlusion was observed following 8% (2/26) of treatments at final angiographic follow-up comprising one case of 10% stenosis and one case of complete PED occlusion.

For aneurysms not occluded at 6-month angiography that underwent later follow-up angiography, the rate of further reduction in contrast filling between 6-month and subsequent angiographic studies was 80% (4/5) for M1 aneurysms and 0% (0/3) for MCA bifurcation aneurysms. Two aneurysms were re-treated with PED following initial PED treatment, with subsequent complete occlusion in one aneurysm and subsequent subtotal filling in the other.

Nineteen major side branches were covered by PED, of which 67% (14/21) remained patent, 24% (5/21) developed stenosis with reduced flow and 10% (2/21) demonstrated complete occlusion with no flow at final angiographic follow-up (Table 3). There were no clinical complications related to coverage of any major side branches.

Table 3.

Outcomes of major side branches covered by PED. Rates reported as n (%).

Major side branch	Total	Patent	Stenotic	Occluded	Clinical consequence
M2	12	8 (67%)	2 (17%)	2 (17%)	0 (0%)
A1	2	0 (0%)	2 (100%)	0 (0%)	0 (0%)
Anterior temporal artery	6	5 (83%)	1 (17%)	0 (0%)	0 (0%)
Posterior communicating artery	1	1 (100%)	0 (0%)	0 (0%)	0 (0%)
Total	21	14 (67%)	5 (24%)	2 (10%)	0 (0%)

Discussion

In this work, we report the safety and efficacy of flow diversion for M1 and MCA bifurcation aneurysms using PED in a multi-center patient cohort. We observed complete aneurysm occlusion in 71% of aneurysms treated and volume reduction in 96%. In M1 segment aneurysms, complete occlusion was observed in 75% of aneurysms and volume reduction was observed in 100% of aneurysms. In MCA bifurcation aneurysms, complete occlusion was observed in 63% of aneurysms and volume reduction was observed in 88% of aneurysms. Technical complications occurred in 7% of all procedures, none of which resulted in clinical sequelae. Non-disabling ischemic complications were observed in 8% of patients. In-stent stenosis or occlusion was observed following 8% of treatments. Covered side branches that developed stenosis or occlusion were clinically silent in all cases.

As flow diverting stents become an increasingly popular method of aneurysm treatment, sustained study of their performance is necessary to provide neurointerventional physicians with guidance for off-label applications. Flow diverting stents facilitate arterial remodeling, leading to aneurysm closure and occlusion, and have demonstrated reasonable performance in a variety of cerebrovascular locations.^26–28 While flow diverting stents have some advantages over conventional treatments, their use in the proximal MCA—where lenticulostriate arteries, the ipsilateral anterior cerebral artery (ACA), or M2 may be covered by the PED—incurs potential risk of occlusion to functionally important arteries. The decision to use flow diversion in this region is further complicated by the availability of other treatment modalities. For instance, clipping demonstrated similar safety compared to endovascular treatment for MCA aneurysms in the ISAT trial. ²⁹ However, a more recent meta-analysis comparing clipping to modern endovascular techniques for the treatment of MCA aneurysms reported better functional outcomes for patients undergoing endovascular treatment. ³⁰ In fact, MCA aneurysms are considered on-label for some new endovascular devices such as the Woven Endobridge Device (WEB) (Microvention, Aliso Viejo, CA). ³¹ , ³² There are also newer flow diverters such as Silk Vista (Balt Extrusion, Montmorency, France), FRED (Microvention), and Surpass (Stryker, Fremont, CA), some of which can be used in conjunction with techniques such as Flow-T stenting. ³³ , ³⁴ The proliferation of viable treatment options for MCA aneurysms should raise the bar for flow diversion in the proximal MCA.

Previous studies describing clinical and angiographic outcomes in MCA aneurysms treated with flow diversion have included predominantly aneurysms at the MCA bifurcation, which often carry unfavorable anatomical features such as being wide-necked and incorporating side-branch vessels. ³⁵ Some studies have demonstrated favorable occlusion rates and safety profiles at the MCA bifurcation, while others have reported worse aneurysm occlusion rates and clinical outcomes. Notably, Briganti et al and Caroff et al report ischemic complications in 27% and 43% of cases, respectively. Our cohort was predominantly composed of M1 segment aneurysms, but also included 8 MCA bifurcation aneurysms. Treated MCA bifurcation aneurysms in our cohort demonstrated more favorable rates of complete occlusion (63%) and volume reduction (88%) than earlier studies, and no patients from this group experienced ischemic or hemorrhagic complications following treatment with PED.

Treatment of M1 aneurysms with PED is not as well described as the treatment of MCA bifurcation aneurysms. These M1 aneurysms may have distinct behavior compared to MCA bifurcation aneurysms given the possible lack of brisk flow in covered side branches that is invariably present in bifurcation aneurysms. Some smaller case series report favorable occlusion rates for PED-treated aneurysms located in the M1. In our series, which includes 17 M1 aneurysms, we observed complete aneurysm occlusion in 75% of aneurysms and volume reduction in 100%. We also observed an ischemic or hemorrhagic complication rate of 13% in this location, which may be reasonable in some scenarios but demands careful consideration of alternative treatment strategies.

Coverage of side branches has understandably prompted apprehension about the risk of infarction, ³⁶ , ³⁷ but other studies have reported a relatively low rate of clinical sequelae. ³⁸ , ³⁹ Both ACAs covered in our cohort were stenotic at angiographic follow-up without associated clinical consequence, and 83% of anterior temporal arteries covered remained completely patent at follow-up imaging. In our cohort, 12 M2 arteries were covered by PED, and at final angiographic follow-up 17% were completely occluded, 17% were stenotic, and 67% remained completely patent. No clinical complications associated with side branch or perforator coverage were observed in our cohort, in line with other studies emphasizing a low rate of these complications in bifurcation aneurysms ³⁷ , ⁴⁰ However, 19% of patients with available cross-sectional imaging had clinically silent infarcts presumably due to perforator coverage. The long-term implications of silent infarcts requires further study, ⁴¹ , ⁴² though the use of high-porosity flow-diverting stents may dampen concerns relating to coverage of both perforators and side vessels by preserving blood flow. ⁴³

Timing of angiographic follow-up may have influenced our findings, particularly considering recent reports of an 11% increase in occlusion rate from 6-month to 24-month follow-up imaging in aneurysms treated with PED. ⁴⁴ In our cohort, 80% of M1 aneurysms that did not occlude by 6 months and underwent later angiographic imaging demonstrated further reductions in contrast filling. A similar tendency was not observed in MCA bifurcation aneurysms. While this observation is based on a small number of patients, it may reflect differences in the kinetics of vascular remodeling in different regions of the proximal MCA. These kinetic differences may also contribute to the higher rates of stenosis or occlusion seen in A1 and M2 branches covered by PED.

The major limitation to this work is that aneurysms included in this study were non-randomly selected for endovascular flow diversion following consideration of alternative open or endovascular treatments by multi-disciplinary experts, or in some cases, failure of these alternative treatments. As such, our cohort is not representative of all M1 and MCA bifurcation aneurysms, and thus our data should not be regarded as a direct comparison with clipping or other endovascular therapies. However, the aneurysms in our cohort highlight the important applications of PED in challenging clinical scenarios where alternative therapies are unattractive. Further limitations in our study arise from its retrospective design, heterogeneity in PED delivery systems employed, variable follow-up times, inclusion of previously-treated aneurysms, and inclusion of a small number of fusiform aneurysms. ⁴⁵ Additionally, while the number of patients in our study is large enough to reach broad conclusions about safety and efficacy, it is insufficient for high precision estimates.

Conclusion

Pipeline embolization of cerebral aneurysms in the M1 segment and MCA bifurcation demonstrates a 71% rate of complete aneurysm occlusion and an 8% rate of ischemic or hemorrhagic complications. Further detailed study is needed to validate these results and better define clinical scenarios in which Pipeline embolization of proximal MCA aneurysms may be reasonable.