Long-term follow-up results of the SMART coil in the endovascular treatment of intracranial aneurysms

Abstract

Background

The Penumbra SMART coil is a novel device that becomes progressively softer from its distal to proximal end to maximize coil packing and prevent microcatheter prolapse or coil migration. Here, we report a large series of patients detailing the long-term experience of a single institution using the SMART coil among patients with intracranial aneurysms (IAs).

Methods

Prospective data of 105 consecutive patients with 106 IAs treated using SMART coils was collected between March 2015 and July 2018. Clinical and angiographic data were analyzed.

Results

Forty-nine patients (46.7%) presented with subarachnoid hemorrhage and 16 (14.2%) had recurrent aneurysms. Two patients had minor intraprocedural ruptures and remained neurologically stable. One patient had a thromboembolic complication with progressive neurologic decline. There was only one case microcatheter prolapse related to placement of a stent before coiling. An initial post-treatment modified Raymond-Roy Occlusion Classification (mRROC) I or II closure was achieved in 56 (52.8%) aneurysms. The average time to last follow-up was 8.4 months at which 70 (81.4%) aneurysms had mRROC I or II occlusion and a major recurrence was seen in 5 (5.8%) patients. Thirteen (12.3%) aneurysms required re-treatment of which one aneurysm was clipped.

Conclusions

The Penumbra SMART coil is safe and effective for the endovascular treatment of appropriately selected IAs. Additional studies at multiple centers comparing safety and efficacy profile over long-term periods to other mainstream coils are necessary.

Introduction

The development of detachable coils for endovascular coil embolization to treat unruptured and ruptured intracranial aneurysms (IAs) has led to great improvements in patient outcomes.^1--3 Assistance devices including balloon catheters,^4-8 implantable stents^9-13 and flow-diverters^14,15 have expanded the range of aneurysms that may be treated endovascularly. Over the last decade, coil technology has continued to vastly improve.⁴ Notable among these changes is significant advancements in delivery systems, improving deployment and detachment of coils, and in the coil implant materials themselves. However, endovascular treatment of intracranial aneurysms (IAs) continues to be limited by incomplete occlusion, recanalization and the need for retreatment.^2,3 Overall recurrence and retreatment rates following coil embolization have been reported as 20.8% and 10.3%, respectively.⁴

Standard platinum coils have a uniform stiffness creating a resistance in the last stage of coil packing for IAs necessitating a further push that can result in coil^5–8 or microcatheter prolapse.^9,10 Microcatheter prolapse is often corrected by repositioning the microcatheter, although a forceful push of the microcatheter may result in penetration of the wall of the aneurysm. To avoid this complication and to maximize coil packing density, the design of the SMART coil (Penumbra Inc., Alameda, CA) was developed. The SMART coil, which was introduced to the market in August, 2015, becomes progressively softer as the coil is deployed allowing for easier deployment with a more stable microcatheter position. The material of the SMART coil is also stretch resistant, mitigating coil unraveling and fracture.

Use of SMART coils for IA treatment has been performed at a limited number of centers in the United States and there have been few studies assessing the safety and efficacy of this new technology. The few case series reporting on the use of SMART coils have shown promising safety and angiographic outcomes including one previous report¹¹ of 17 IAs treated using SMART coils in which 12 achieved immediate RROC I or II occlusion and there was one case of microcatheter prolapse related to placement of a stent prior to coiling. Here, we report 106 consecutive patients treated using SMART coils and also discuss long-term outcomes. This is the first study describing a large cohort of IA patients treated with the SMART coil with long-term clinical and angiographic outcomes.

Materials and methods

Data collection

We prospectively collected data from medical records of patients with intracranial aneurysms (IAs) treated using at least one SMART coil during endovascular embolization at a single academic center. Institutional Review Board approval was obtained for this study. Patient consent was not obtained due to minimal risk of this study. The study sample was collected by reviewing the prospectively maintained neuroendovascular database during the period between August 2015 and July 2018. Patient demographic data collected included age, gender, history of smoking, hypertension and family history of aneurysms. Angiographic data extracted were location, maximum diameter, neck width, presence of blebs or dome irregularities and whether the aneurysm treated was ruptured or recurrent. Procedural details recorded were total number of coils deployed per patient, number of SMART coils deployed, and the use of adjunctive coiling techniques such as balloon-assisted, stent-assisted or flow-diversion. Procedural complications included microcatheter prolapse, intraprocedural rupture and thromboembolic complications. Outcomes were recorded immediately after embolization and at 6 months follow up by degree of aneurysm occlusion and recurrent aneurysm requiring retreatment. Degree of aneurysm occlusion was classified using the modified Raymond-Roy Occlusion Classification (mRROC)¹² as follows: class I, complete occlusion; class II, residual neck; class IIIa, contrast opacification within the coil interstices of a residual aneurysm; class IIIb, contrast opacification outside the coil interstices, along the residual aneurysm wall. mRROC classes I and II were considered good angiographic outcomes as these classes have reported favorable outcomes.^12–14

Device details

In this series, all aneurysms were treated with at least one SMART coil. The Neuron 070 6 F guiding catheter (Penumbra Inc, Alameda, CA) with an Excelsior SL-10 (Boston Scientific, Natick, MA) microcatheter tip was advanced into the aneurysm for coil delivery. Stent-assisted coiling was performed with the Low-Profile Intraluminal Support Device (LVIS Jr., MicroVention, Tustin, CA) and flow-diversion with adjunctive coiling was performed with the Pipeline Embolization Device (Medtronic; Minneapolis, MN). The choice of coil sizing remains the same as that of other non-SMART coils relying on judgement based on tactile feedback, catheter positioning and filling of the aneurysm. Non-SMART coils were used at the discretion of the treating interventionalist, most frequently in the setting of unavailable SMART coil sizes.

Statistical analysis

Descriptive statistics were calculated for all variables reported. Categorical variables are reported as frequencies. The mean was reported as a measure of central tendency for all continuous variables. No comparative statistics were conducted. All analyses were performed using Prism (GraphPad Software, Inc.) version 8.

Results

Patient and aneurysm characteristics

One-hundred and five patients with 106 intracranial aneurysms (IAs) were treated with SMART coils. Table 1 summarizes the patient demographics and aneurysm characteristics. Of the 105 patients, the mean age was 60.0 (range 12–90) and 74 (70.5%) were female. Risk factors identified included history of smoking, hypertension and family history of IAs among 53 (50.5%), 52 (49.5%) and 12 (11.4%) patients, respectively. The most common aneurysm location was the anterior communicating artery (ACoA) (33.0%), followed by internal carotid artery (ICA) (15.1%), posterior communicating artery (PCoA) (15.1%) and basilar artery apex (14.2%). The mean aneurysm maximum diameter was 7.3 mm (range 1.9–18.1), with 84.0% of aneurysms less than 10 mm. The mean neck size was 3.7 mm (range 1.5–8.9, n = 86). Forty-nine (46.7%) aneurysms were ruptured upon presentation and 16 (14.2%) aneurysms treated with SMART coils were recurrent. Twelve (11.3%) aneurysms have blebs or dome irregularities. One patient had a bilobed aneurysm and each lobe was coiled separately so each lobe was counted as a separate aneurysm in this series (Figure 1).

Table 1.

Patient demographics and procedural details of 106 intracranial aneurysms treated with Penumbra SMART coils.

Characteristic	No. (%) or mean ± SD (range)
Patient data
Total patients	105
Total aneurysms treated	106
Age (years)	60.0 ± 14.6 (12–90)
Female	74 (70.5)
History of smoking	53 (50.5)
History of hypertension	52 (49.5)
Family history of intracranial aneurysm(s)	12 (11.4)
Aneurysm data
Location
ACoA	35 (33.0)
ICA	16 (15.1)
PCoA	16 (15.1)
BA apex	15 (14.2)
MCA	9 (8.5)
ACA	4 (3.8)
PICA	3 (2.8)
OphA	3 (2.8)
PCA	3 (2.8)
SCA	2 (1.9)
Aneurysm size (mm)	7.3 mm ± 3.8 (1.9–18.1)
Small (<10 mm)	89 (84.0)
Large (10–25 mm)	16 (15.1)
Giant (>25 mm)	1 (0.94)
Neck size (mm)	3.7 mm ± 1.9 (1.5–8.9)a
Ruptured	49 (46.7)
Dissecting	1 (0.94)
Recurrent	16 (14.2)
Blebs or dome irregularity	12 (11.3)
Embolization
Aneurysms treated with only SMART coils	86
Total number of coils per aneurysm	5.0 ± 2.5 (1–13)
Number of SMART coils per aneurysm	4.4 ± 2.5 (1–12)
Stand-alone coiling	68 (64.2)
Stent-assisted coiling	26 (24.5)
Flow-diversionb with coiling	10 (9.4)
Balloon-assisted coiling	2 (0.94)
Complications
Microcatheter prolapse	1c
Intraprocedural rupture	2 (1.9)
Thromboembolic complications	1 (0.94)

ACoA: anterior communicating artery; ICA: internal carotid artery; PCoA: posterior communicating artery; BA: basilar artery; MCA: middle cerebral artery; ACA: anterior cerebral artery; PICA: posterior inferior cerebellar artery; OphA: ophthalamic artery; PCA: posterior cerebral artery; SCA: superior cerebellar artery.

^aMean out of 86 total aneurysms with recorded neck sizes.

^bPipeline embolization device.

^cOne case of microcatheter prolapse related to placement of a stent before coiling.

Figure 1.

(a) Ruptured bilobed anterior communicating artery aneurysm prior to coil embolization. (b) Coil embolization of right lobe using SMART coils. (c) Left lobe of aneurysm embolized. M2-M3 middle cerebral artery (MCA) perfusion defect during treatment of aneurysm. (d) Subsequent contrast injection showing improved perfusion of M2-M3 MCA after injection of intravenous heparin.

Procedural details and complications

Table 1 also reports the procedural details and complications among this study cohort. The mean number of total coils and SMART coils used per patient were 5.0 (range 1–13) and 4.4 (range 1–12), respectively. Non-SMART coils were used to treat 20 (18.9%) aneurysms. The average number of non-SMART coils used among these 20 aneurysms was 3.4. Sixty-eight (64.2%) aneurysms were treated using coiling alone, 26 (24.5%) with stent-assisted coiling, 10 (9.4%) with flow diversion with adjuvant coiling and two (1.9%) patients with balloon-assisted coiling. Microcatheter prolapse occurred in one case during placement of a stent before any coiling was performed. There were three other complications including two (1.9%) intraprocedural aneurysm ruptures and one (0.94%) thromboembolic complication. The two intraprocedural aneurysm ruptures occurred among a right supraclinoid ICA aneurysm and a left ACA A1/A2 junction aneurysm. The right supraclinoid ICA aneurysm had ruptured during the early loop of the first coil. Several measures were taken to secure the aneurysm including balloon inflation, reversal of anticoagulation with protamine, rapid continued coiling with 3 additional aneurysms and administration of mannitol (once the aneurysm had been secured). The left ACA A1/A2 junction aneurysm had a very small rupture with minimal contrast extravasation secondary to a small coil loop extruding through the aneurysm wall. Hemostasis was rapidly achieved with continued coiling and administration of protamine.

One patient had presented with subarachnoid hemorrhage (SAH) and coil embolization was complicated by re-rupture with coil extrusion of the first SMART coil through the wall of the aneurysm. Several measures were taken to secure the aneurysm and minimize hemorrhage including balloon inflation, reversal of anticoagulation with protamine, administration of mannitol and rapid continued coiling with an additional 3 SMART coils. A subsequent CT was performed revealing no dramatic changes in extent of hemorrhage or ventricle size from pre-procedural imaging. The patient remained stable throughout the rest of the hospital course and was discharged home without evidence of neurologic deficit. Another patient had minor extravasation during deployment of the last coil during a stent-assisted embolization for an unruptured aneurysm. This patient was discharged home without evidence of clinical sequelae on post-operative day 2.

The third patient experienced an intraoperative thromboembolic complication in which a left M2-M3 middle cerebral artery (MCA) branch occlusion caused a parenchymal defect. The degree of occlusion improved over subsequent contrast injections after administration of 4000 units of intravenous heparin. Given the dynamic nature of the occlusion following heparin administration and the distal position of the occlusion, a decision was made to not proceed with mechanical thrombectomy. In the postoperative setting. there was also a right A2 anterior cerebral artery (ACA) occlusion causing parenchymal defect. The patient continued to decline neurologically with an expanding area of MCA infarct evident on CT imaging and intracranial pressure spikes to the 50’s. On post-operative day 1, the patient was made comfort measures only and died on post-operative day 2.

Initial and long-term angiographic outcomes

Table 2 lists the immediate angiographic outcomes of all IAs in this series. Immediate mRROC class I or II occlusion was present among 52.8% of aneurysms. Of these aneurysms, 69.6% were coiled without any assisting devices, 14.3% were stent-assisted, 14.3% were flow-diverted with adjuvant coiling and 1.8% were balloon-assisted. There were twenty-seven (25.5%) and twenty-three (21.7%) aneurysms classified as mRROC IIIa and mRROC IIIb, respectively. While most of these aneurysms were coiled without assisting devices (63.0% and 52.2% for mRROC IIIa and IIIb aneurysms, respectively), a larger portion of these aneurysms were stent assisted than in the mRROC I/II groups (37.0% and 34.8% for mRROC IIIa and IIIb aneurysms, respectively).

Table 2.

Immediate angiographic outcomes of 106 intracranial aneurysms treated with SMART coils.

Treatment type	mRROC I/II	mRROC IIIa	mRROC IIIb
All (n = 106)	56 (52.8%)	27 (25.5%)	23 (21.7%)
Stand-alone coiling (n = 68)	39 (57.3%)	17 (25.0%)	12 (17.6%)
Stent-assisted coiling (n = 26)	8 (30.8%)	10 (38.4%)	8 (30.8%)
Flow-diversiona with coiling	8 (80%)	0	2 (20%)
Balloon-assisted coiling	1 (50%)	0	1 (50%)

mRROC: modified Raymond-Roy Occlusion Classification.

^aPipeline embolization device.

Table 3 describes both the short-term and long-term angiographic outcomes of this series. The average length of time to first angiographic follow-up was 5.9 months. There were 17 patients without angiographic follow-up including 7 deceased patients and 10 lost to follow-up. At 6 months after treatment, mRROC class I or II occlusion was achieved in 80.0% of aneurysms. A minor recurrence, defined as a change in mRROC from I to II, occurred in 2 (2.5%) aneurysms and a major recurrence, defined as a change in mRROC from I or II to III, was evident among 5 (6.3%) aneurysms. Mean time to last angiographic follow-up was 8.4 months (range 0.3–34). Seventy patients (81.4%, n = 86) had mRROC class I or II occlusion at the last angiographic follow-up. There were 5 (5.8%) cases of aneurysm progression from mRROC class I/II immediately after treatment to III at last follow-up. Re-treatment was performed among 13 (12.3%) aneurysms. The average time from initial to re-treatment was 316 days (range 17–622). Flow diversion with adjuvant coiling was the most common method of re-treatment (46.2%) followed by stent-assisted coiling (30.8%), stand-alone coiling (15.4%) and one (7.7%) case of clipping.

Table 3.

Short and long-term angiographic outcomes and aneurysms treated with SMART coils requiring re-treatment.

Characteristic	No. (%) or mean ± SD (range)
6 months follow up (n = 80)
mRROC I/II	64 (80.0)
mRROC IIIa	5 (6.3)
mRROC IIIb	11 (13.8)
Minor recurrence: mRROC I to II	2 (2.5%)
Major recurrence: mRROC I/II to IIIa/b	5 (6.3%)
Time to last follow-up	8.8 months ± 6.9 (0.3–34)
Last follow-up (n = 86)
mRROC I/II	70 (81.4)
mRROC IIIa	8 (9.3)
mRROC IIIb	8 (9.3)
Minor recurrence: mRROC I to II	0
Major recurrence: mRROC I/II to IIIa/b	5 (5.8)
Re-treatment
Total number requiring re-treatment	13 (12.3)
Time to re-treatment	316 days ± 208 (17–622)
Stand-alone coiling	2 (15.4a)
Stent-assisted coiling	4 (30.8a)
Flow diversionb with adjuvant coiling	6 (46.2a)
Clipping	1 (7.7a)

mRROC: modified Raymond Roy Occlusion Classification.

^aPercentage out of n = 13 total re-treated aneurysms.

^bPipeline embolization device.

Discussion

This is the largest series reporting outcomes after intracranial aneurysm (IA) embolization with SMART coils and also reports long-term angiographic outcomes. We reported a series of 106 aneurysms treated with SMART coils over the course of 3 years. A wide range of IAs were described in this series. Many aneurysms had wide necks with an average neck size of 3.7 mm and nearly half (46.7%) of patients presented with ruptured aneurysms. Of the 106 aneurysms treated, post-procedure angiographic modified Raymond-Roy Occlusion Classification (mRROC) grade I or II¹² was obtained in 52.8%, although at 6 months and at last angiographic follow-up the proportions of mRROC grade I or II aneurysms were 80.0% and 81.4%, respectively, indicating good long-term outcomes. A large proportion of aneurysms with initial mRROC grade IIIa or IIIb closure were stent-assisted, flow-diverted or recurrent. There were few complications. One patient suffered a thromboembolic complication while 2 patients had minor intraprocedural ruptures with no clinical sequelae.

Five smaller series describing initial outcomes using SMART coils for IAs have been reported.^11,15–18 The first series was an initial report¹¹ of 17 aneurysms in which there was one case of microcatheter prolapse in the setting of mechanical limitations imposed by a stent and mRROC I or II occlusions was achieved in 12 aneurysms. Ilyas et al.¹⁷ reported another series of 32 patients with 33 aneurysms treated using SMART coils without any instances of microcatheter prolapse, intraprocedural aneurysm rupture or thromboembolism although there was one case of device malfunction. RROC grade I or II was achieved in 48% and 27% of the 33 aneurysms. Spiotta et al.¹⁶ described 59 patients with SMART coils, 44% of which were treated in the setting of subarachnoid hemorrhage. RROC grade I occlusion was in 33.9% and II in 37.3% of aneurysms. Daniel et al.¹⁵ reported another SMART coil series consisting of 49 patients in which SMART coils were solely used in 96% of cases. This study also showed a high mRROC I or II occlusion rate at 6 months (90%) with few complications (one case of microcatheter prolapse and 4 cases of thromboembolic events). Sokolwski et al.¹⁸ most recently described a series of 33 patients with 34 aneurysms. mRROC grades I, II, IIIa and IIIb were achieved in 62%, 26%, 3% and 9%, respectively.

There are several technological advancements incorporated into the SMART coil allowing for improved aneurysmal sac filling in addition to prevention of aneurysm perforation and microcatheter prolapse. The SMART coil creates tight, complex shapes within the aneurysmal sac and the coil becomes progressively softer along its length, unlike traditional microcoils which are of uniform stiffness. The distal segment of a coil may cause trauma to the aneurysm wall, increasing the risk of intraprocedural rupture. The leading tip of the SMART coil has an extrasoft 2-dimensional loop which may mitigate the trauma associated with entry into the aneurysm. Microcatheter stability is another important feature of this coiling system. As the proximal segment of a standard coil is deployed, catheter prolapse may occur due to the growing force exerted by an increasingly dense coil mass; however, the SMART coil, as previously described, becomes softer along its length with diminishing force required and smooth coil deployment. In this series, we did not encounter any instances of microcatheter prolapse during coiling. There was only one case of microcatheter prolapse during stent placement before any coils were deployed.

The outcomes reported here using the SMART coil are similar to the widely used platinum coils and even the few novel coils recently developed and reported in prospective studies. The Hydrogel Endovascular Aneurysm Treatment Trial (HEAT) described outcomes of 600 patients randomized to coiling with either bare platinum coils or hydrogel coated coils.¹⁹ The incidences of major recurrences (defined as progression of RROC from 1 or 2 to 3) were 12.8% and 20.7% for the HydroCoil Embolic System (HES) and the bare platinum coils (BPCs), respectively. The incidences of major recurrences in this series was significantly smaller (5.8%). However, retreatment rates in the HEAT trial were 5.2% and 8.3% with HES and BPCs which are lower than described in the current series (12.3%). The Barricade coil system (Blockade/Balt, Montmorency, France) is a novel bare platinum coil that is electrically detached and becomes softer as the coils get smaller. Long-term (1 year) outcomes using Barricade coils were recently published reporting retreatment in 10 (9.7%) patients out of a series of 132 patients which is similar to the rate of retreatment in this series (13.2%).²⁰

There were several limitations to this study. First this was a single-center case series with the selection, treatment and referral biases inherent to a study performed at one center. Another significant limitation to this study was non-SMART coils were used to supplement the embolization when specific SMART coil sizes were not available. While non-SMART coils were used for only 18.9% of cases, results of this study may not precisely reflect the safety and long-term angiographic outcomes with the use of SMART coils solely. Future long-term studies at multiple centers directly comparing outcomes using SMART coils with other mainstream coils are necessary to further understand the clinical impact of this technological advancement.

Conclusions

This is a large, retrospective series reporting long-term angiographic outcomes of 106 intracranial aneurysms treated with SMART coils demonstrating safety and long-term efficacy among a wide range of aneurysm locations and sizes in both initial and recurrent as well as ruptured and unruptured settings. Future studies at multiple centers directly comparing the efficacy of the use of SMART coils with other traditional coils are necessary to further clarify the role of this device in endovascular treatment of intracranial aneurysms.