Is a small coil more effective as a finishing coil for the embolization of intracranial aneurysms? Outcomes using GALAXY G3™ MINI micro-coils as finishing coils in a single center

Abstract

Objective

The purpose of this study was to compare the outcomes of coil embolization using a 0.009 inches primary outer diameter coil as finishing coil (FC) to that of 0.01 inches.

Methods

From February and August 2020, 131 aneurysms that performed coil embolization using FC with a second loop diameter of 1 mm, were reviewed retrospectively, conducting propensity score matching and logistic regression analysis. Angiographic results such as, occlusion grade, packing density, failure and event were compared between 0.009 inches coil of GALAXY G3™ MINI microcoil (n = 54) and 0.01 inches coils (n = 77).

Results

There were no statistically significant differences between two groups, but more events occurred in the 0.009 group. (Odds ratio, 3.65; 95% CI, 1.06-12.55; P = 0.031) In the results of coil embolization, successful occlusion occlusion (complete occlusion and residual neck) was identified more in the 0.01 group. After propensity score matching, the variables in each group were similar, but the successful occlusion was higher in the 0.01 group as in the total population. Events tended to occur more frequently in the 0.009 inch group, and logistic regression analysis showed slightly higher events in the angled microcatheter. (48.3% versus 76.9%., P = 0.075), Also, the 0.009 inch FC is an independent risk factor. (Odds ratio, 3.84; 95% CI, 1.07-13.80; P = 0.039)

Conclusions

Using 0.01 inches coils as FC increased the packing density after the procedure, and showed more successful occlusion than using a 0.009 inches coil. The probability of unexpected events was observed more than three times in the 0.009 inch group.

Introduction

Endovascular coil embolization is currently the main treatment for an intracranial aneurysm, along with surgical clipping. For successful treatment and to prevent aneurysm recurrence, it is important to select an appropriate coil and fill it as tightly as possible.^1–3 Generally the endovascular coil embolization involves three distinct stages: 1) formation of a basket or framework for the subsequent deposition of the coil, 2) filling of the aneurysm volume, and 3) the finishing stage of the coil packing, which maximizes attenuation. Smaller and softer coils are usually selected at this final stage of coil embolization, and these very soft coils are often referred to as finishing coils (FCs).^4–6

Recently, many companies introduced various types of FCs with unique characteristics such as shape, stiffness, and size to improve coil embolization and achieve successful outcomes. The GALAXY G3™ MINI micro-coil (GM) (CERENOVUS, Irvine, CA, USA), the smallest coil with a primary outer diameter (POD) of 0.009 inches, was recently released and is widely used as an FC. Since FCs are commonly used in complicated situations where they are placed into a previously placed coil mass gap and inserted into a space that is disproportionately smaller than the default setting of the coil, the smaller the primary outer diameter, the softer the coil, so higher packing density and effective coil embolization can be expected. ⁷ However, no studies have compared the effectiveness of the smaller GM coil as an FC to other coils with 0.01-inch PODs. In this study, for the first time, we compared the outcomes of coil embolization using 0.009-inch GMs to those of other coils [Target 360 nano (Stryker, Fremont, CA, USA) and Microplex hypersoft 3 D (Mirovention Terumo, Tustin, CA, USA)] of 0.01 inches and considered various factors affecting outcomes of the FC.

Materials and methods

Study population and data collection

From February to August 2020, a total of 125 patients with 146 intracranial aneurysms underwent coil embolization at our single institution. Among them, coil embolization using FCs with a second loop diameter of 1 mm was performed in 131 aneurysms. Patients treated for non-saccular type aneurysms such as oncotic, dissecting, and blood blister-like aneurysms, the use of FCs larger than 1 mm in diameter, or those with unexpected complications during the procedure were excluded. Retrospectively, we reviewed medical records to assess patient-related factors such as age, gender, hypertension, diabetes mellitus, hyperlipidemia, smoking history, subarachnoid hemorrhage (SAH), aneurysm location, and the surgical records were reviewed to identify operation-related factors such as the number and type of coils, the shapes of the microcatheters, intracranial stent use, and other unexpected events during the procedure. This study was approved by the Institutional Review Board of the author’s institution.

Endovascular procedures

In most cases, diagnostic angiography and three-dimensional rotational image reconstruction were undertaken to determine the shapes and sizes (width, depth, and neck) of the aneurysms using an Artis Q biplane or an Artis zee biplane (Siemens Healthineers, Erlangen, Germany). All coil embolization procedures were performed under general anesthesia, and a bolus intravenous injection of unfractionated heparin (2500–3000IU) was given after placing the femoral sheaths and an additional dose (1000 IU) was administered hourly while monitoring the activated clotting times, except for in the rupture cases. Four skilled endovascular neurosurgeons (SU.K, J.H, WJ.J, and HG.L) certified by the Korean Neuroendovascular Society (KoNES) performed appropriate treatment for each cerebral aneurysm and tried to fill the aneurysmal sac as densely as possible until one more coil could not be placed.

Angiographic results and clinical outcomes

The final angiographic results after coil embolization were classified as complete occlusion, residual neck, or residual sac using the Roy Raymond scale. Both complete occlusion and residual neck classifications represent successful occlusion. ⁸ The choice of FC and the types of microcatheter, and whether to treat aneurysms by stent-assisting coiling was at the discretion of the operator. It was considered to fail when an attempt to fill the FC was unsuccessful, and an event was considered to have occurred when an unintended situation such as a coil loop intruded out of the aneurysm sac or the microcatheter dropped out while inserting the FC (Figure 1). The packing density (PD) was calculated using the AngioCals app based on the shape and size of the aneurysm, as well as the type and size of the coils, and measured before FC insertion and after the procedure was completed. The microcatheter used to insert the FC was classified as an angled mcirocatheter when pre-shaped more than 90 degrees or steam-shaped into various forms.

Figure 1.

Before (a) and after (b) the final coil (FC) was inserted into the coil embolization of a 65-year-old woman. It was a successful coil embolization that filled the remaining space of the aneurysm neck (black arrow). However, before (c) and after (d) the FC was inserted into the coil embolization of a 78-year-old woman, the loop of the FC protruded into the parent vessel, and before (e) and after (f) the FC was inserted into the coil embolization of a 77-year-old women, the microcatheter unexpectedly came out while inserting the FC, so these were classified as events.

Statistical analysis

The statistical analyses were performed using SPSS version 25.0 (IBM Corporation, Armonk, NY, USA). The continuous variables are presented as means and standard deviations and compared between the two subgroups using Fisher’s exact or the chi-squared test, as appropriate, and categorical variables were analyzed using the Student’s t-test. Propensity score matching was performed to compensate for the group-wise imbalances in the baseline characteristics that could potentially skew the FC filling results. Score-matching for each patient reflecting the PD before FC insertion was performed within R version 4.0.3 (R Foundation for Statistical Computing, Vienna, Austria) using the nearest-neighbor package for optimal 1:1 matching. A probability value of <0.05 was considered statistically significant. Odds ratios (OR) and their 95% confidence intervals (CIs) were also reported.

Results

Baseline characteristics of the patients and aneurysms

A total of 131 aneurysms were selected for this study. The mean age of the study population was 63.8 ± 10.6 years (range, 28 to 85 years) and most of the patients (101/131, 77.1%) were female. The mean volume of the aneurysms was 29.77 ± 29.08 mm³ (range, 2.23 to 389.3). The clinical parameters were distributed as follows: hypertension, 71 patients (56.5%); diabetes mellitus, 25 (19.8%); dyslipidemia 26(19.8%); smoking history 16 (12.2%), and SAH 10 (7.6%). Successful occlusion (complete occlusion or residual neck) after immediate coil embolization was achieved in 99 cases (75.6%), with a mean packing density of 37.1 ± 13.9%. The baseline characteristics of the patients and treated aneurysms are presented in Table 1.

Table 1.

Angiographic and clinical characteristics before and after propensity score matching in coil embolization using different finishing coils.

Variable	Total population				Propensity score matching (n = 108)			Odds or HR (95% CI)
	(n = 131)	0.01-inch FC (n = 77)	0.009-inch FC (n = 54)	P value	0.01-inch FC (n = 54)	0.009-inch FC (n = 54)	P value
Age (years)	63.8 ± 10.6	62.6 ± 11.7	65.8 ± 7.8	0.059	61.9 ± 7.8	65.8 ± 7.8	0.051
Female	101 (71.1%)	58 (75.3%)	43(79.6%)	0.564	42 (77.8%)	43(79.6%)	0.814
HTN	71 (56.5%)	44 (57.1%)	30 (55.6%)	0.857	30 (55.6%)	30 (55.6%)	1.000
DM	25 (19.1%)	11 (14.3%)	14 (25.9%)	0.095	7 (12.9%)	14 (25.9%)	0.088
Dyslipidemia	26 (19.8%)	17 (22.1%)	9 (16.7%)	0.445	12 (22.2%)	9 (16.7%)	0.466
Smoking	16 (12.2%)	10 (13.0%)	6 (11.1%)	0.747	8 (14.8%)	6 (11.1%)	0.567
SAH	10 (7.6%)	4 (5.2%)	6 (11.1%)	0.509	4 (7.4%)	6 (11.1%)	0.507
Location				0.427			0.422
ACA	15 (11.5%)	6 (7.8%)	9 (16.7%)		4 (7.4%)	9 (16.7%)
MCA	27 (20.6%)	18 (23.4%)	9 (16.7%)		12 (22.2%)	9 (16.7%)
ICA	84 (64.1%)	50 (64.9%)	34 (63.0%)		37 (68.5%)	34 (63.0%)
BA, VA	5 (3.8%)	3 (3.9%)	2 (3.7%)		1 (1.9%)	2 (3.7%)
Aneurysm volume (mm3)	30.7 ± 44.1	29.8 ± 29.1	32.0 ± 59.6	0.799	29.6 ± 25.3	32.0 ± 59.6	0.781
Angled microcatheter	67 (51.1%)	42 (54.5%)	25 (46.3%)	0.353	27 (50.0%)	25 (46.3%)	0.700
Stent assisted	98 (74.8%)	56 (72.7%)	42 (77.8%)	0.512	39 (72.2%)	42 (77.8%)	0.505
Coil length 1 mm	68 (51.9%)	36 (46.8%)	32(59.3%)	0.158	32 (59.3%)	32(59.3%)	1.000
PD before FC (%) (A)	32.4 ± 11.9	30.3 ± 10.2	35.4 ± 12.9	0.020	33.3 ± 10.9	35.4 ± 12.9	0.359
PD after FC (%) (B)	37.1 ± 13.9	36.4 ± 13.3	38.1 ± 14.8	0.503	38.8 ± 13.7	38.1 ± 14.8	0.800
B-A	4.7 ± 6.0	6.1 ± 7.0	2.6 ± 3.3	0.000	5.5 ± 6.4	2.6 ± 3.3	0.005
Successful occlusion	99 (75.6%)	63 (81.8%)	36 (66.7%)	0.047	45 (83.3%)	36 (66.7%)	0.046	0.44 (0.20–0.99)
Fail	34 (26.0%)	18 (23.4%)	16 (29.6%)	0.422	15 (27.8%)	16 (29.6%)	0.832
Event	13 (9.9%)	4 (5.2%)	9 (16.7%)	0.031	3 (5.6%)	9 (16.7%)	0.066	3.65 (1.06–12.55)

HR Hazzard ratio, CI Confidence interval, FC Finishing coil, HTN Hypertension, DM Diabetes mellitus, SAH Subarachnoid hemorrhage, ACA Aanterior cerebral artery, MCA Middle cerebral artery, ICA Internal carotid artery, BA Basilar artery, VA Vertebral artery, PD Packing density.The statistically significant values are presented in bold (P < 0.005).

Comparison of coil embolization outcomes using different finishing coils

To evaluate the effect of FC size on coil embolization, GM (n = 54) with a POD of 0.009 inches was compared to two coils (n = 77, Target nano: 49 and Microplex hypersoft: 28) with the same 0.01-inch size with a secondary loop of 1 mm. There were no statistically significant differences between the two groups in the volume and location of the aneurysms, as well as clinical factors and procedural factors, such as the use of angled catheters, stent use, and the length of the FC. The PD before placing the FC with smaller size, but not after placing the PC, in the GM group was significantly higher (P = 0.020) and the number of events also increased. (OR = 3.65; 95% CI: 1.06–12.55; P = 0.031). Even though there was not significant difference in the failure to insert the FC, immediate successful occlusion by coil embolization was identified more often in the larger FC group (P = 0.047). After 1:1 propensity score matching of the subjects, the groups were reclassified into 54 cases of 0.009-inch coils and 54 cases of 0.01-inch coils. (Target nano: 28 and Microplex hypersoft: 26). The characteristics and angiographic variables in each group were similar, and the PD before FC was not significantly different between the two groups, but the number of successful occlusions was higher in the 0.001-inch group like in the total population (P = 0.046). And, although not statistically significant, events tended to occur more frequently in the 0.009-inch group (P = 0.066) (Table 1).

As shown in Table 2, univariate and multivariate logistic regression analysis of the event risk factors showed no association with other factors, but a somewhat higher event rate was found for the angled microcatheter (48.3% versus 76.9%, P = 0.075). In addition, the 0.009- inch GM FC was an independent risk factor of events (OR = 3.84; 95% CI: 1.07–13.80; P = 0.039).

Table 2.

Univariate and multivariate logistic regression analysis of event risk factors during coil embolization in all study participants.

Variable	Event (%)		Univariate analysis	Multivariate analysis	OR (95% CI)
	No (n = 118)	Yes (n = 13)	P value	P value	P value
Age > 65 (years)	54 (45.8%)	6 (46.2%)	0.560
Female	91 (77.1%)	10 (76.9%)	0.908
HTN	67 (56.8%)	7 (53.8%)	0.237
DM	24 (20.3%)	1 (7.7 %)	0.064
Dyslipidemia	22 (18.6%)	4 (30.8%)	0.221
Smoking	14 (11.9%)	2 (15.4%)	0.466
SAH	10 (8.5%)	0 (0%)	0.999
Location			0.767
ACA	14 (11.9%)	1 (7.7 %)
MCA	26 (22.0%)	1 (7.7%)
ICA	73 (61.9%)	11 (84.6%)
BA, VA	5 (4.2%)	0 (0%)
Aneurysm volume >30 (mm3)	34 (28.8%)	2 (15.4 %)	0.691
Angled microcatheter	57 (48.3%)	10 (76.9%)	0.075	0.107
Stent assisted	85 (72.0%)	13 (100%)	0.998	0.998
Coil length 1 mm	60 (50.8%)	8 (61.5%)	0.898
PD before FC > 30(%) (A)	59 (60.4%)	8 (61.5 %)	0.298
Successful occlusion	90 (76.3%)	9 (69.2%)	0.912
Fail	31 (26.3%)	3 (23.1%)	0.510
0.009-inch FC	45 (38.1%)	9 (69.2%)	0.028	0.039	3.84 (1.07–13.80)

OR Odds ratio, CI Confidence interval, HTN Hypertension, DM Diabetes mellitus, SAH Subarachnoid hemorrhage, ACA Aanterior cerebral artery, MCA Middle cerebral artery, ICA Internal carotid artery, BA Basilar artery, VA Vertebral artery, PD Packing density, FC Finishing coil.The statistically significant values are presented in bold (P < 0.005).

Discussion

This was the first study to compare the clinical outcomes of using two different POD sizes as FCs. A few papers have been published on the results of experimental measurements of coils in the in vitro environment, or clinical outcomes using a single specific coil as the FC.^4,7,9 Although comparative studies of different types of stents have been performed,^10,11 there are few comparative analyses of clinical results according to different coil characteristics.

The FC is usually preferred to be very soft, and the coil softness is affected by the size of the structures composing the coil. In general, a smaller POD was expected to be softer and produce more effective clinical results, but the result of this study did not show that. Stiffness (or softness) is defined as the spring constant of a secondary structure because the coil structure is similar in appearance to a spring, and the theoretical coil stiffness coefficient (K) can be calculated as follows:^12,13

$$\mathrm{K}=\frac{{D1}^{4}G}{8{D2}^{3}n};\mathrm{\hspace{1em}}\mathit{Stiffness}\propto \frac{{D1}^4}{{D2}^3}$$

where D1 is the coil wire diameter, G is the shear modulus of the coil wire, D2 is the POD, and n is the number of turns per unit distance. The smaller the value of K, the softer the coil. Stiffness is directly proportional to the fourth power of D1 and inversely proportional to the third power of D2. Therefore, the smaller thd POD, the higher the value of K, and the coil became less soft. In fact, the specifications of Target 360 nano and Hypersoft 3 D, the 0.01-inch coils used in this study, are 0.00125 inches for D1 and 0.0100 for D2, and the K factor (D1⁴/D2³) is calculated as 2.441 × 10⁻⁶. In contrast, a 0.009-inch GM coil has a D1 of 0.00120, a D2 of 0.009, and a K factor estimated to be 2.844 × 10⁻⁶, which is expected to be stiffer with higher values.

Higher PD is associated with lower recanalization aneurysm rates and is an important factor for stable, complete occlusion.^3,14,15 As the FC is considered to have a large effect on the PD after the procedure, we compared the results before and after FC insertion in this study. In the total population, the 0.009-inch FC group had a significantly higher PD before FC (30.3 ± 10.2 versus 35.4 ± 12.9, P = 0.020), but propensity score matching case-control analysis was used to increase the comparability of the two groups. The demographic and angiographic variables of both groups became similar after 1:1 propensity score matching, and the PD before FC was also not statistically significant (33.3 ± 10.9 versus 35.4 ± 12.9, P = 0.359). However, the difference in PD before and after FC and the numbers of successful occlusion were significantly higher in the 0.01-inch FC group (P = 0.005, and P = 0.046, respectively). Similar to a report by Zhang et al. on the effect of a first large-sized coil, ¹⁶ if there is no significant difference in the failure of FC insertions, as in our study, using a larger-sized POD coil as the FC could result in a high PD and increase the successful occlusion rate. ¹⁷

The softness of the coil is subjectively evaluated in various ways depending upon the operator, but events such as microcatheter dropout or coil loop herniation were reviewed as more objective outcome data. Fortunately, there were no thromboembolic complications in this study, but events occurred more than three times more often in the 0.009-inch group, and logistic regression analysis showed that the GMs with a small POD had almost a 4-fold higher risk of event occurrence (Tables 1 and 2). This phenomenon is thought to be caused by the law of action and reaction when the coil and delivery wire are inserted into the aneurysm sac, and as reported in other experimental studies, the microcatheter in an unstable state is kicked back and the coil tail straightens.^4,18,19 As shown in Figure 2, when 1 mm × 2 cm FCs of the same size used in this study were inserted into a pre-shaped 90-degree microcatheter, the tip of the microcatheter was moved differently for each coil, about 10 degrees for Hypersoft 3 D, 25 degrees for Target nano, and 30 degrees for GM, respectively. In particular, considering that the tip of the catheter shifts as the coil almost exits the microcatheter, in addition to coil size, other factors such as the thickness of the delivery wire and the detach area characteristics seem to affect the movement of the catheter tip (Supplementary video).^20,21 The above experimental environment was different from the actual situation where the catheter tip is fixed in the aneurysm sac, but the results of this study suggest that relatively more events occurred due to tip movement in the angled catheter, (48.3% versus 76.9%, P = 0.075) and more events occurred with the GM, the coil with the largest angle of catheter tip change.

Figure 2.

Comparison of the movement of the microcatheter tip during coil insertion. The microcatheter used was an Excelsior SL-10 pre-shaped 90 (Stryker, Kalamazoo, MI, USA): (a) Control, (b) Target 360 nano 1 mm × 2 cm (Stryker), (c) Microplex hypersoft 3 D 1 mm × 2 cm (Mirovention Terumo, Tustin, CA, USA), and (d) GALAXY G3TM MINI micro-coil 1 mm × 2 cm (CERENOVUS, Irvine, CA, USA).

There were several limitations to this study. First, patient data was collected for a short period of seven months in a single institution and analyzed retrospectively, and clinical results were that only packing density and occlusion grade were assessed immediately after procedure and a result of the long term follow-up was insufficient. Furthermore, there were also biases, such as different skill levels of the operators, differences in the coil embolization technique, and preferences in coil selection. Among the study variables, whether a coil insertion failure or event occurred, and the grade of aneurysm occlusion were determined through discussion with the authors, but it could not be completely objective and free from errors. Although statistically corrected using propensity score matching, it seems that there was a selection bias that preferred a small coil when PD was relatively full. In addition, as the two 0.001-inch Target nano and Microplex hypersoft coils were included in the same group, the impact of the different properties of the two coils, such as shape, the softness of the delivery wire, materials and mechanism of the coil-wire junction, and balance, on the results may have been overlooked.

Conclusion

Using 0.01-inch coils as FCs increased the PD after the procedure, and showed more successful occlusion, compared to using 0.009-inch GMs. When a smaller size coil was used as the FC, the probability of unexpected events increased more than three-fold. When choosing the FC, not just small size should be considered but factors such as the property of coil compositions and the interaction with the microcatheter.