Triple Aspiration versus Conventional Aspiration Techniques: A Randomized In Vitro Evaluation

Triple-Aspiration Technique and Control Groups

We tested the performance of 3 different aspiration techniques: triple aspiration, single-syringe aspiration, and vacuum pump aspiration.

For the triple-aspiration technique, a 3-way stopcock was connected to a 60-mL aspiration syringe (Vaclok; Merit Medical Systems) and an aspiration catheter (6F Sofia; MicroVention). With the stopcock open to both the 60-mL syringe and the aspiration catheter, the syringe was fully pulled back, and the first aspiration was performed. Then, the stopcock was immediately closed to the aspiration catheter, and any air within the syringe was expelled through the open side port of the 3-way stopcock. Next, the system was closed to the sideport, and a second aspiration was performed using the 60-mL syringe. Then, by means of the same process, any remaining air was expelled through the side port, and subsequently, the third aspiration was performed. We waited for at least 2 minutes after the third aspiration to ensure standardization across all treatment arms, and then the aspiration catheter was removed with the syringe remaining locked. The technique is also demonstrated in a video (Online Supplemental Data).

The single-syringe aspiration was also performed using the 60-mL Vaclok syringe. For the pump aspiration, we used the Riptide vacuum pump (Medtronic). The Riptide pump was set to the highest vacuum pressure achievable after the pressure gauge integrated in the pump reached the plateau, and aspiration was continued for at least 2 minutes for both techniques.

3D-Printed Neurovascular Model

The manufacturing process of the 3D-printed in vitro thrombectomy model has been described by Li et al.¹² Briefly, the head and neck arterial vasculature of patients with acute large-vessel occlusion was extracted from anonymized CT angiographies and refined for 3D-printing using the Meshmixer software (Autodesk). Then, the model was 3D-printed using a commercially available resin and printer (Elastic 50A, Form 3 SLA Printer; Formlabs). The elastic 50A resin was specifically chosen because it exhibits a Young Modulus of 1.7 megapascal, mimicking the elasticity of arterial vasculature.

The 3D-printed model had bifurcation MCA anatomy bilaterally and patent anterior (AcomA) and posterior communicating (PcomA) arteries. It included an aortic arch, bilateral carotid arteries (ICA), MCAs (up to 2 distal M2 MCA branches), anterior cerebral arteries (up to the proximal A2 segment), AcomA, PcomA, and posterior cerebral arteries (up to the proximal P2 posterior cerebral artery segments). To mimic transfemoral access, we connected a silicone tubing to the aortic arch, and an 8F introducer sheath was placed into the silicone tubing. Additionally, 2 filters with a 50- and 70-µm pore size were placed at the inflow and outflow of the model for particle filtering and collection purposes, respectively. The 3D-printed model and thrombectomy setup are summarized in Fig 1A.

• Download figure
• Open in new tab
• Download powerpoint

FIG 1.

Experimental thrombectomy setup (A) and aspiration pressure curves (B).

Clot Analogs and Thrombectomy Setup

The performances of aspiration thrombectomy techniques were tested with soft (red blood cell [RBC-rich]) and stiff (fibrin-rich) clot analogs. The RBC and fresh frozen plasma donations were obtained from the Mayo Clinic Blood Transfusion Service. The RBC-rich clots consisted of 16-mL RBC, 32-mL fresh frozen plasma, and 2 mL of 5% calcium chloride (Sigma-Aldrich), and fibrin-rich clots consisted of 47.5 mL of fresh frozen plasma, 0.5 mL of RBC, and 2 mL of 5% calcium chloride. Clot volumes and diameters were standardized to maintain consistency among the replicates. Blood components were put into 50-mL tubes and were mixed by inversion. Then, clot analogs were formed overnight at room temperature. The next day, all clot analogs were cut in 30-mm segments using a 3D-printed rectangular prism with a lumen to ensure standardization across replicates. Subsequently, the clot analogs were measured using a scale, and those significantly smaller or larger than 3 cm were excluded. Next, clot analogs were lodged into the terminus of the right ICA in dynamic flow conditions. Then, clots were measured again, and any outliers were discarded. Fragmented clots and clots that have embolized beyond the terminal ICA were removed from the circulation and not included in the analysis. Following proper occlusion, experiments were randomized into 1 of the 3 treatment arms (ratio 1:1:1), ensuring an equal distribution of stiff and soft clots (ratio 1:1) within each arm. After embolization, the lengths of the clots inside the model were also recorded, and the normal distribution of clot lengths across groups was ensured.

A roller pump was used to circulate approximately 800 mL/min of 37°C water within the model. A 6F FUBUKI guiding catheter (Asahi Intecc Medical) was placed in the right proximal ICA. Then, an aspiration catheter (6F Sofia) was advanced to the occlusion site with the help of a 0.027-inch microcatheter and 0.014-inch micro guidewire. Clots were not passed with the microcatheter or the Micro Guidewire during catheter navigation to avoid introducing distal emboli to the system. After reaching the occlusion site, the microcatheter and micro guidewire were removed, and contract aspiration was performed for at least 2 minutes as described above. Thirty-four replicates (17 RBC-rich and 17 fibrin-rich) were performed for each aspiration technique.

Technical Outcomes, Vacuum Pressures, and Analysis of Distal Emboli

Study end points were complete clot removal, first-pass efficacy (FPE), the number of thrombectomy attempts, vacuum pressure, and distal embolization. Because microscopic clot fragments might be missed during the experiment, complete clot removal was considered analogous to a modified TICI (mTICI) scale score of ≥2c. Because the distinction between mTICI 2a and 2b may not be reliable in an in vitro model, complete clot removal was chosen as the primary recanalization end point. Up to 3 thrombectomy attempts were performed to achieve complete clot removal. FPE was defined as the ability of the aspiration technique to achieve complete clot removal after the first attempt (mTICI 2c/3).

To measure the generated vacuum pressures, we connected the aspiration tubing of the vacuum pump or the 60-mL syringe to a pressure transducer (Model ASDXRRX015PDAA5; Honeywell) and replicated the aspiration approaches described previously. We acquired 5 measurements for each approach, and the pressure data are presented as the mean (SD).

To evaluate distal embolization, we placed a 70-μm cell strainer at the outflow of the model distal to the intracranial vasculature for each replicate. After each experiment, the cell strainer was removed, and the distal embolization was analyzed using an image-processing algorithm, as described by Li et al.¹² The algorithm provided the following data: the total count of emboli of >70 μm, the Feret diameter of the largest embolus, the mean Feret diameter of the emboli, the total count of emboli of >1 mm, and the total area of emboli.

Statistics

Statistical analyses were conducted with a statistical software package (SPSS, Version 25.0 for Windows; IBM). The Shapiro-Wilk test was used to examine the distribution of the data. A 1-way ANOVA test and Tukey post hoc analysis were used to examine the differences in normally distributed continuous data. The Kruskal-Wallis test was used for non-normally distributed continuous data. The χ² and Fisher-Freeman-Halton tests were used for categoric data collected in the benchtop thrombectomy model. For all statistical tests, a P value < .05 was considered statistically significant.