The impact of microcatheter-to-vessel size ratio on distal embolization during mechanical thrombectomy—an in vitro quantitative study

Abstract

Background and purpose

Mechanical thrombectomy (MT) is the gold standard for treating large vessel occlusions. Given the variability in vessel anatomy among patients and the sometimes arbitrary selection of devices by neurointerventionalists, the choice of microcatheter size requires thorough evaluation. We aim to evaluate the impact of device-to-vessel size ratio on periprocedural distal embolization.

Materials and methods

Fragment-prone clot analogs (length = 9.86±0.07 mm) were used to embolize three different vessels (lumen = 2.0, 2.5, 3.5 mm) in a three-dimensional-printed neurovascular model. Three different microcatheter sizes (0.017″, 0.021″, 0.027″) were used to cross the lesion and subsequently, migrated clot fragments were collected in an outflow filter for image analysis. Experiments were conducted both with and without a microguidewire. A total of 180 experiments were performed: 60 for each M1 segment of middle cerebral artery size, including 20 for each microcatheter—10 with J-shaped microguidewire and 10 without.

Results

Across all vessels, the 0.027″ microcatheter caused more distal embolizations compared to 0.017″ (p = 0.04) and 0.021″ (p = 0.01). In the 2 mm M1-MCA, 0.017″ microcatheter reduced emboli compared to 0.021″ (p = 0.062) and 0.027″ (p = 0.017). Procedures in the 2 mm vessel are significantly more prone to embolization compared to larger M1 segments (p < 0.05). Microcatheter-to-vessel ratio ≥0.38 significantly increases risk of distal embolization. The use of microguidewire in the procedures did not have any impact on distal embolization (p = 0.871).

Conclusion

A larger device-to-vessel size ratio induces an increase in distal embolization. Neurointerventionalists should carefully consider vessel anatomy for appropriate microcatheter size selection to minimize the risk of distal embolization.

Introduction

Multiple randomized trials ¹ have demonstrated the benefits of mechanical thrombectomy (MT) in patients with acute ischemic stroke due to large vessel occlusions, consistently associating perfect final reperfusion grades with better functional outcomes. However, the variability in vessel anatomy among patients represents an important challenge when it comes to selecting appropriate MT devices by interventionalists.

Several studies have characterized distal emboli generated during MT,^2–4 highlighting the possible influence on final reperfusion outcomes. Distal embolization can limit the extent of recanalization, which may lead to suboptimal clinical outcomes. In this context, minimizing this risk is critical to improve final reperfusion scores. A previous study ⁵ explored the impact of different microcatheter diameters on distal emboli generation when crossing the clot, showing that the use of smaller devices causes smaller and fewer particles.

As a rule, the smallest microcatheter should be used to cross the occluding clots. However, smaller catheters compromise the size of the retrieving devices deliverable through them. Therefore, a compromise between microcatheter size and affordable potentially generated emboli should be made for each clot location or arterial size. While the general effect of microcatheter size has been studied, the emboli generated according to the device-to-vessel size ratio remain unexplored. Given the large variability of the vessel diameter at the level of the treated occlusion, ⁶ the impact of the microcatheter size may be completely different in each case.

In this context, we aim to investigate the potential effect of the device-to-vessel size ratio on the generation of distal emboli to provide valuable insight that will allow to optimize device selection during MT.

Methods

Neurovascular 3D-printed model manufacturing procedure

The neurovascular model (FLOWCAT, Barcelona, Spain) was based on the vascular anatomies extracted from anonymized computed tomography angiography (CTA) images. The manufacturing procedure consists of medical image segmentation, where a preliminary 3D geometry of the vascular anatomy is generated, followed by mesh modeling to simplify the anatomy and prepare a printable model, 3D-printing and postprinting processing, and model assembly and setting in a flow-loop.

3DSlicer^7,8 and Autodesk Meshmixer 3.5 were used in model segmentation and design. Preform software and Form 3 SLA Printer (Formlabs, Inc., Somerville, MA, USA) were used for 3D printing.

The neurovascular phantom comprises the aortic arch, bilateral carotid arteries, middle cerebral arteries (MCA) (up to 2 distal M2-MCA branches), anterior cerebral arteries (up to proximal A2), anterior communicating artery, posterior communicating arteries, and posterior cerebral arteries (up to proximal P2-PCA segments). Three models with different M1-MCA were created to represent the lower (2.0 mm), intermediate (2.5 mm), and higher range (3.5 mm) of M1 segment's diameter variability.

Friable clot analogs

The process involving the fabrication of analog clots has been previously reported. ¹⁰ Briefly, reagents are mixed in the following order: acrylamide 40% (2.5 mL), bis-acrylamide 2% (1.125 mL), ammonium persulfate 10% (0.1 mL), tetramethylethylenediamine (TEMED, 0.01 mL) (reagents from Bio-Rad, CA, USA), and distilled water (6.375 mL).

Silicone tubings (2.0, 2.5, and 3.5 mm) were attached to a syringe to withdraw the liquid gel content, where a thrombus analog is formed in approximately 15 minutes. Before each experiment, the clot analog was cut into smaller thrombi (length = 9.92±0.78 mm), and embolized into the M1-MCA of the neurovascular phantom. These clot analogs are characterized by high stiffness (elasticity tangent modulus, E_t = 86.3 ± 3.8 kPa), high friction (µ_s = 0.54 ± 0.11), and friability to represent worst-case scenarios in the neurovascular model, mimicking refractory occlusions with low thrombus-device engagement and a high risk of distal embolization.

Detailed procedure

An experimental setup (Figure 1) was designed to evaluate distal emboli associated with microcatheter navigation through the clot. The setup includes a 3D-printed neurovascular model connected to a flow-loop at 37°C supplied with a flow rate of 800 mL/min. Clot analogs (length: 9.86±0.07 mm, diameters: 1.86±0.09 mm; 2.37±0.12 mm; 3.4±0.19 mm) were inserted into the model to simulate M1-MCA segment occlusions. Additionally, a 100-µm pore size filter was connected to the outflow of the model so that particles larger than pore size could be collected.

Figure 1.

Experimental setup.

Each experiment started with the placement of a guiding catheter (Neuronmax, Penumbra inc, CA, USA) at the C2 segment level of the internal carotid artery followed by the navigation of a distal access catheter (React 71, Medtronic, MN) up to the proximal M1-MCA segment. Three microcatheters (Neuroslider 17, 21, 27, Acandis, DE) were then advanced through the clot.

Two treatment arms were established to compare outcomes by navigating the microcatheter through the clot with a J-shaped microguidewire (pORTAL 0.014″, Phenox, DE) or without it. ⁹ When using a J-shaped microwire, the wire was first navigated through the clot and the microcatheters were then advanced over the wire. Afterwards, the devices were fully retrieved and the outflow filter was subjected to particle analysis. In experiments without a microwire, it was removed before crossing the clot and only the microcatheter was advanced through the occluding lesion.

A total of 180 clots were crossed: in each of the three M1-MCA segments (diameters: 2, 2.5, and 3.5 mm) and 20 replicates with each of the three microcatheters (with microwire n = 10 / without microwire n = 10).

Device-to-vessel size ratio

To compute the microcatheter-to-vessel size ratio, outer diameters of each microcatheter (Neuroslider 17 OD: 0.65 mm; 21 OD: 0.75 mm; 27 OD: 1 mm) were divided by the corresponding vessel diameter (2.0, 2.5, and 3.5 mm).

Emboli characterization

An image processing algorithm was developed on MATLAB R2024b (MathWorks, Inc., Natick, MA, USA) to analyze the distal emboli and provide a size distribution frequency diagram (online Supplemental material 1). The program starts with the recognition of an RGB (Red, Green, Blue) image taken by a high-resolution digital camera (IPEVO, Inc., Sunnyvale, CA, USA). A binarization is applied, where the emboli are highlighted (white) and the background is removed (black). For further processing and accurate measurement of desired metrics, an 8 mm diameter circle was placed in the image as a reference to calculate the pixel–microns relationship. The algorithm provided the area of each particle in the filter to obtain the parameters of the largest embolus across all particles in the filter (max area) and the total area formed by all emboli (total area).

Statistical analysis

Results were expressed as mean ± SD. Data was analyzed using SciPy library from Python. Welch's t-test was used to calculate pairwise comparisons. A negative binomial regression model was computed to determine the relationship between microcatheter size, vessel size, and emboli area parameters.

Results

A total of 180 experiments with equal representation across the three microcatheters (0.017″, 0.021″, and 0.027″), three M1-MCA vessel diameters (2.0, 2.5, and 3.5 mm) and use of a J-shaped microguidewire were performed. The maximum embolus area (max area) and the total area formed by emboli (total area) collected in the outflow filter were analyzed.

Microguidewire influence

The use of microguidewire did not significantly impact thrombus fragmentation and migration. Overall, the emboli generated with the presence of a microwire (max area: 0.78±1.93 mm²; total area: 1.08±2.09 mm²) did not yield any significant difference compared to particles generated in the experiments without a microwire (max area: 0.84±2.41 mm², p = 0.872; total area: 1.14±2.79 mm², p = 0.871). Moreover, the impact of the presence of the guidewire was similar in all vessel diameters, and all microcatheter sizes (online Supplemental material 2).

Impact of microcatheter size

The impact of microcatheter size is graphically represented in Figure 2. In general, the larger the microcatheter diameter the larger the emboli total area. The 0.027″ microcatheter consistently shows higher values in both emboli area parameters (max area: 1.54±3.3 mm²; total area: 1.88±3.73 mm²) than the 0.017″ microcatheter (max area: 0.31±1.0 mm², p = 0.01; total area: 0.51±1.05 mm², p = 0.01) and the 0.021″ microcatheter (max area: 0.57±1.26 mm², p = 0.04; total area: 0.95±1.59 mm², p = 0.08). The 0.021″ microcatheter was associated with non significantly larger areas of emboli compared to the 0.017″ microcatheter (max area: p = 0.22; total area: p = 0.08).

Figure 2.

Graphical representation of emboli generated by the clot-crossing maneuver with each microcatheter in each M1-MCA model. Distal embolization in terms of total emboli area (left) and largest embolus area (center). Table summarizing mean and standard deviation of each device and vessel size combination (right).

Influence of vessel size

The influence of vessel size is graphically represented in Figure 2. The analysis showed increased total emboli area in the 2.0 mm vessel (2.02±3.57 mm²) as compared to the larger vessel sizes (2.5 mm: 0.73±1.9 mm², p = 0.045; 3.5 mm: 0.59±0.98 mm², p = 0.012). No significant difference was seen between 2.5 and 3.5 mm M1-MCA models (p = 0.491). Similarly, the 2.0 mm vessel yielded significantly higher max area (1.70±3.3 mm²) than any other M1-MCA model (2.5 mm: 0.34±1.26 mm², p = 0.012; 3.5 mm: 0.38±0.92 mm², p = 0.016). The 2.5 mm still produced non significantly larger max areas than the 3.5 mm vessel (p = 1.000).

In 2.0 mm M1-MCA, the 0.017″ microcatheter (max area: 0.47±1.68 mm²; total area: 0.57±1.69 mm²) generated non significantly smaller areas than the 0.021″ microcatheter (max area: 1.28±2.01 mm², p = 0.176; total area: 1.85±2.45 mm², p = 0.062), whereas compared to the 0.027″ microcatheter, both area parameters revealed significantly higher values for the larger microcatheter (max area: 3.36±4.71 mm², p = 0.016; total area: 3.64±5.06 mm², p = 0.017). The 0.027″ microcatheter did not yield significantly higher emboli areas than the 0.021″ microcatheter (max area: p = 0.081, total area: p = 0.166).

In 2.5 and 3.5 mm vessel segments, no significant differences were found.

Microcatheter to vessel size ratio

The relationship between the microcatheter-to-vessel size ratio and distal embolization, represented by max area and total area, shows a rising trend as the ratio increases (Figure 3(a)). Specifically, the average max area values are significantly smaller for ratios below 0.32 (0.33±0.95 mm²) compared to the ratios above 0.38 (1.76±3.33 mm², p = 0.002). Similarly, the average total area values were significantly higher for ratios above 0.38 (2.23±3.79 mm²) compared to lower ratios (0.55±0.99 mm², p = 0.001; Figure 3(b)).

Figure 3.

Distal embolization according to microcatheter-to-vessel size ratio in terms of maximum embolus area and total emboli area. (a) All ratios for all microcatheters and (b) grouped according to ratio ≤0.32 versus ≥0.38. (**p < 0.01).

Microcatheter and vessel size as predictors of embolization areas

To further assess the impact of vessel size on distal embolization, we computed two negative binomial regression models using as references the potential worst combination of microcatheter and vessel diameter (0.027″ device and 2 mm vessel), and potential best scenario (0.017″ device and 3.5 mm vessel).

The model's predictors explain 23.56% and 32.53% of the variance in total area and max area, respectively. Table 1 shows incident rate ratios (IRRs), 95% confidence intervals (CI) and corresponding p-value for the model's predictors. For example, an IRR of 0.31 indicates that the specific class of the independent variable (microcatheter 0.017″ crossing 2-mm M1-MCA occlusion) is associated with a 69% decrease in the dependent variable (total emboli area) with respect to the reference class of the independent variable (microcatheter 0.027″ crossing 2-mm M1-MCA occlusion).

Table 1.

IRR, CI and p-value coefficients for the negative binomial regression models. (A) shows coefficients for microcatheter and vessel sizes when compared to the corresponding reference sizes (microcatheter 0.027’’ and M1-MCA of 2mm). (B) shows coefficients for microcatheter and vessel sizes when compared to the corresponding reference sizes (microcatheter 0.017’’ and M1-MCA of 3.5mm).

A	Reference classes: microcatheter = 0.027'’ and M1-MCA size = 2mm
	Microcatheter size/vessel size	IRR	CI lower	CI upper	p-value
Total area	Microcatheter 0.017''	0.31	0.18	0.53	<0.001
	Microcatheter 0.021''	0.51	0.31	0.84	0.008
	M1-MCA 2.5mm	0.38	0.23	0.64	<0.001
	M1-MCA 3.5mm	0.32	0.19	0.55	<0.001
Maximum area	Microcatheter 0.017''	0.23	0.12	0.43	<0.001
	Microcatheter 0.021''	0.37	0.21	0.66	0.001
	M1-MCA 2.5mm	0.21	0.11	0.38	<0.001
	M1-MCA 3.5mm	0.25	0.13	0.45	<0.001
B	Reference classes: microcatheter = 0.017'’ and M1-MCA size = 3.5mm
	Microcatheter size/vessel size	IRR	CI Lower	CI Upper	p-value
Total Area	Microcatheter 0.027''	3.24	1.88	5.59	<0.001
	Microcatheter 0.021''	1.65	0.93	2.94	0.089
	M1-MCA 2.5mm	1.18	0.66	2.1	0.582
	M1-MCA 2.0mm	3.09	1.82	5.24	<0.001
Maximum area	Microcatheter 0.027''	4.39	2.32	8.3	<0.001
	Microcatheter 0.021''	1.63	0.81	3.28	0.168
	M1-MCA 2.5mm	0.84	0.41	1.72	0.634
	M1-MCA 2.0mm	4.08	2.24	7.42	<0.001

CI: confidence interval; IRR: incident rate ratio; MCA: middle cerebral arteries.

Discussion

Our study provides valuable insights into the risk of distal embolization when crossing an occlusion with a microcatheter in the M1-MCA territory. The results evidence that both microcatheter and vessel size play a crucial role in possible endovascular complications, while the use of a microguidewire does not.

As stated above and in line with a previous study, larger microcatheter sizes lead to larger emboli areas, most notably in the 0.027″ microcatheter. Due to the microcatheter-to-vessel size ratio, this phenomenon was expected given the larger cross-sectional area of this device which will more likely fragment or even displace the thrombus, with an exponential effect in smaller vessels. Whilst the 0.021″ catheter generated smaller emboli areas than the 0.027″ microcatheter, it still carried a bigger risk in distal embolization than the smaller device chosen for this study, confirming the findings of Gounis et al. ⁵ on favoring smaller microcatheter sizes when possible.

In terms of vessel size, smaller vessels were associated with significantly higher emboli areas compared to larger vessels, suggesting that smaller vessels indeed allow for more microcatheter-to-vessel interactions, leading to higher risk of thrombus fragmentation. Given the clinical importance of the M1-MCA territory, where larger-vessel occlusions often occur, ¹¹ the choice of microcatheter size becomes even more critical as the 2.0 mm vessel diameter is a common size for this vascular territory. ¹²

In general, bibliographical evidence^6,12,13 shows variability across cerebrovascular anterior circulation geometry associating a wide range of diameters to different vessel segments. Thus, our study can be extrapolated to other clinically significant scenarios like the internal carotid artery (ICA) or the M2 segment of the MCA or even more distal. This variability reinforces the need to carefully consider vessel size when selecting microcatheter sizes. Larger microcatheters carry a greater risk in smaller vessels, but this impact diminishes with increasing vessel diameter. For instance, as seen in Figure 2, crossing a 3.5 mm diameter ICA ⁶ occlusion with a 0.027″ microcatheter, induces higher embolization risks compared to smaller catheters, but the relative impact is less critical in larger vessels than in narrower ones. Similarly, while microguidewire use did not significantly affect results, the potential for larger-caliber macroguidewires to generate distal emboli through an unfavorable device-to-vessel size ratio remains unexplored and should be addressed in future studies.

In fact, our study identifies similar device-to-vessel size ratios across different catheter and vessel sizes, which appear to correlate with comparable embolization risks. Specific ratios of 32.5% (microcatheter 0.017″ and 2 mm vessel), 30% (microcatheter 0.021″ and 2.5 mm vessel), and 28.5% (microcatheter 0.027″ and 3.5 mm vessel) all produce nearly equivalent risks of distal embolization. This suggests that if this ratio is within a certain range, the risk of embolization may not vary critically. Cerebrovascular geometry becomes indeed a critical factor in balancing the choice of microcatheter size to minimize risk of distal embolization.

The negative binomial regression model applied in our analysis for both reference classes scenarios provided a robust fit for our data. In the first scenario (reference: 0.027″ microcatheter and 2.0 mm vessel), smaller microcatheters and larger vessel sizes were significantly associated with decreased embolization areas. Specifically, IRR coefficients show the superior safety profile of the 0.017″ microcatheter, with bigger vessels amplifying this preventive effect. On the other hand, in the second scenario (reference: 0.017″ microcatheter and 3.5 mm vessel), the largest microcatheter (0.027″) and the smallest vessel (2 mm) consistently exhibited a notorious increase in embolization areas, confirming the findings of the previous regression model. Notably, intermediate microcatheter and vessel sizes were not significantly associated with an increase in embolization areas, suggesting that the impact may lie between extremes of smaller and larger device-to-vessel size ratios. In summary, our models underscore the importance of precise device selection that needs to be carefully addressed to specific vascular geometries to prevent distal embolization during MT.

Limitations of the study

In vitro studies have inherent limitations, especially in replicating clot-vessel interactions, as well as the heterogeneous composition of thrombi and their mechanical properties. Our findings were obtained in a controlled setup with continuous water perfusion. Our results might differ under more realistic physiological conditions.

While our study tried maintaining device uniformity throughout all the experiments, results might differ with other microcatheter and microguidewire fabricants.

In this study, images are only acquired from a single angle and hence the particles can only be examined in two dimensions. Therefore, particle volume is completely neglected in this study when it has been suggested to be determinant. ¹⁴ An improved setup to acquire 3D images would enhance our study and further confirm our results.

Conclusion

A higher incidence of distal embolization has been shown to be related to a higher device-to-vessel size ratio. Thus, it is important that neurointerventionalists consider the anatomy of the target vessel for a safer and more adequate microcatheter selection. When smaller microcatheters are deemed feasible for a procedure, the risk of distal embolization is minimized and ultimately helps improve procedural outcomes.