Results

In this study, a two-stage study design consisting of in vitro experiments with anatomically accurate human aortic arch models and in vivo large animal experiments was adopted to systematically evaluate the core preclinical performance of the integrated system, and all experiments were successfully completed. In vitro experiments verified the embolic capture efficacy of the device and its puncture accuracy under types Ⅰ to Ⅲ aortic arch configurations. During in vivo experiments, the animals maintained stable vital signs with no intraoperative adverse events, and the surgical technical success rate was 100%.

In vitro experimental results

Embolus capture efficiency

A paired design was used to evaluate the capture rate for each microsphere size. Capture rate was calculated as: [(Blank control count − experimental count)/(Blank control count) × 100%]. The raw data of each particle size group and the descriptive statistical results of the capture rate are shown in Table II. As the particle size of the fluorescent microspheres increased from 100 to 150 μm, the capture rate showed a significant gradient increase: the mean value was (24.2 ± 8.2)% in the 100 μm group (range, 16.3%-33.3%), (87.1 ± 2.7)% in the 120 μm group (range, 84.2%-89.7%), and reached (98.1 ± 0.9)% in the 150 μm group (range, 97.2%-99.0%). The three groups of data showed small intragroup fluctuation with no outliers.

Microsphere particle size (μm) | No. of microspheres in the blank control group (counts) | No. of microspheres in the experimental group (counts) | Capture efficiency (%)
100 | 96.3 ± 9.6 | 72.7 ± 8.0 | 24.2 ± 8.2
120 | 100.3 ± 3.1 | 13.3 ± 2.5 | 87.1 ± 2.7
150 | 99.7 ± 4.0 | 2.0 ± 1.0 | 98.1 ± 0.9

Normality and homogeneity of variance test

The Shapiro-Wilk test showed that the capture efficiency of the three groups all conformed to normal distribution (100 μm: W = 0.999, P = .942; 120 μm: W = 0.893, P = .363; 150 μm: W = 0.964, P = .637; all P > .05). Homogeneity of variances was verified by Levene test, and the results showed homogeneous variances among the three groups (F = 3.555; P = .092 > .05), which met the prerequisites for parametric test.

Validation of intragroup capture effect (blank control group vs experimental group)

Paired t-test results showed that the capture effect was notable in the 120 μm group (t = 56.955; d__f = 2; P < .001) and the 150 μm group (t = 55.372; d__f = 2; P < .001); a certain capture effect was also observed in the 100 μm group (t = 4.313; d__f = 2; P = .050; Table III).

Microsphere particle size (μm) | t value | Degrees of freedom (df) | P value | 95% CI
100 | 4.313 | 2 | .05 | −0.012-9.512
120 | 56.955 | 2 | <.001 | 83.025-91.175
150 | 55.372 | 2 | <.001 | 96.834-99.366

Comparison of intergroup differences in capture efficiency

One-way analysis of variance showed that capture efficiency differed among the three groups (F = 174.771; d__f = 2; 6, P < .001), with rates increasing as particle size increased. The results of post hoc multiple comparisons using the Tukey honestly significant difference method were as follows: the differences between the 100 μm group and 120 μm group (P < .001), as well as between the 100 μm group and 150 μm group (P < .001), were both extremely statistically significant. There was no statistically significant difference in capture efficiency between the 120 μm group and 150 μm group (Tukey honestly significant difference method, P = .093); only the uncorrected least significant difference method for multiple testing indicated a statistically significant intergroup difference (P = .042). The overall results still confirmed the dose-effect trend that capture efficiency increased with the increase of microsphere particle size (Fig 2, A).

Fig 2: Capture efficiency of the integrated system for embolic microspheres of different particle sizes and correlation analysis. A, Intergroup comparison of capture efficiency for microspheres of different particle sizes. The differences in capture rate between the 120 and 150 μm microsphere groups and the 100 μm microsphere group were extremely statistically significant (∗∗∗∗P < .001); B, Spearman rank correlation analysis between the particle size of fluorescent microspheres and capture efficiency, which showed a strong positive correlation between the two (rs = 0.949, n = 9; P < .001). The experiment was performed using an embolic protection filter with a 120 μm pore size, with three independent repeated experiments conducted for each particle size in each group.

Correlation between particle size and capture efficiency

Spearman rank correlation analysis showed that there was a strong positive correlation between the particle size of fluorescent microspheres and capture efficiency (_r_s = 0.949; n = 9; P < .001), indicating that particle size was a key factor influencing the filter capture efficacy (Fig 2, B).

Performance of active steering directional puncture

The in vitro experiments systematically evaluated the puncture concentration performance of this integrated system for the three main supra-aortic arteries (such as left subclavian artery, left common carotid artery, and brachiocephalic trunk) under three aortic arch configurations (types I-Ⅲ). The results of the core measurement indicator “puncture coverage ratio” (area of the puncture range/projection area of the artery ostium) are shown in Fig 3 and Table IV.

Fig 3: Spatial distribution of puncture sites for the three supra-aortic branch vessels under types Ⅰto Ⅲ aortic arch anatomies, demonstrating the directional accuracy and concentration performance of the active steering puncture mechanism. A-C, Puncture coverage areas of the left subclavian artery under types Ⅰto Ⅲ aortic arches, respectively; D-F, Puncture coverage areas of the left common carotid artery under types Ⅰto Ⅲ aortic arches, respectively; G-I, Puncture coverage areas of the brachiocephalic trunk under types Ⅰto Ⅲ aortic arches, respectively. Red hatched areas represent the geometric projection of the target vessel ostium on the stent graft membrane; green solid areas represent the actual distribution range of all successful puncture sites. Across all nine test scenarios, all puncture sites were concentrated within the projection area of the target vessel ostium with no out-of-bounds puncture events.

Artery type | Aortic arch types | Projection area of artery ostium, mm2 | Area of puncture range, mm2 | Puncture coverage ratio, %
Left subclavian artery | Ⅰ | 70.69 | 48.63 | 68.8
 | Ⅱ | 72.81 | 38.64 | 53.1
 | Ⅲ | 74.93 | 31.42 | 41.9
Left common carotid artery | Ⅰ | 63.62 | 33.34 | 52.4
 | Ⅱ | 62.49 | 29.32 | 46.9
 | Ⅲ | 60.58 | 26.68 | 44.0
Brachiocephalic trunk | Ⅰ | 91.84 | 42.85 | 46.7
 | Ⅱ | 79.74 | 36.81 | 46.2
 | Ⅲ | 48.99 | 28.01 | 57.2

Across all nine test scenarios, the measured values of the puncture coverage ratio ranged from 41.9% to 68.8% (Table IV), indicating that all puncture sites were highly concentrated within the projection area of the target vessel ostium, with no out-of-bounds puncture observed. With the increase in the anatomical complexity of the aortic arch, the puncture concentration of different arteries presented a differentiated trend.

For the left subclavian artery and left common carotid artery, the measured values of the puncture coverage ratio showed a gradual decreasing trend as the arch configuration evolved from type Ⅰ to type Ⅲ. The ratio for the left subclavian artery decreased from 68.8% in the type Ⅰ arch to 41.9% in the type Ⅲ arch, whereas the ratio for the left common carotid artery decreased from 52.4% to 44.0%.

Of particular note were the results for the brachiocephalic trunk, which presented a variation pattern distinct from the two aforementioned arteries: the puncture coverage ratio reached the minimum value (46.2%) under the type Ⅱ arch, whereas under the type Ⅲ arch with the most restricted anatomical space, this ratio increased to 57.2% (Fig 4, A).

Fig 4: Changing trend of effective puncture coverage ratio of supra-aortic branch vessels under different aortic arch configurations. A, Bar graph of the puncture coverage ratio of different branch vessels under types Ⅰ to Ⅲ aortic arches; B, Trend line graph of the puncture coverage ratio changing with the anatomical complexity of the aortic arch. The left subclavian artery and left common carotid artery showed a monotonic decreasing trend, whereas the brachiocephalic trunk presented a “V-shaped” trend of initial decrease followed by rebound. The X-axis represents the aortic arch type, and the Y-axis represents the puncture coverage ratio (%).

Visualization analysis of puncture concentration trend

To intuitively demonstrate the variation trend of the above measured values with the anatomical complexity of the aortic arch, a scatter and line plot of the puncture coverage ratio was generated with arch type (types Ⅰ-Ⅲ) set as the ordinal variable (Fig 4, B).

As shown in Fig 4, the connecting line of the datapoints for the left subclavian artery and left common carotid artery presented a clear negative monotonic trend. In contrast, the connecting line of the data points for the brachiocephalic trunk clearly showed a non-monotonic “V-shaped” pattern: a slight decrease from type Ⅰ to type Ⅱ arch, followed by a marked rebound in the type Ⅲ arch. This visualization result corroborated the aforementioned findings, indicating that under the extreme condition of significant compression of the aortic arch anatomical space (the projection area of the brachiocephalic trunk in the type Ⅲ arch was reduced by approximately 46.7% compared with that in the type Ⅰ arch), the variation pattern of puncture concentration for the brachiocephalic trunk was different from that of the left subclavian artery and left common carotid artery.

Qualitative observation of air embolus capture

This part was a preliminary experiment designed to preliminarily observe the potential capture effect of the embolic protection filter on air embolic. The fully deployed embolic protection filter achieved fragmentation and partial entrapment of injected air bubbles; however, the clinical relevance of this observation requires further investigation in larger, quantitative studies.

In vivo experimental results

General conditions and feasibility endpoint

All six female Ukrainian Large White pigs enrolled in this study successfully completed the full simulated ISF and vascular reconstruction workflow of “embolic protection filter deployment-directional puncture-branch stent implantation,” with no operation failure, device-related complications, or experimental interruption events. The technical success rate in the model reached 100%, which verified the feasibility of the basic operation of the device.

Efficiency endpoint

The statistical results of the core operational efficiency indicators are shown in Table V.

Observation indicator | Mean ± standard deviation | Range
Time from adjustment completion to puncture and fenestration, minutes | 0.15 ± 0.03 | 0.08-0.18
Total operative time, minutes | 38.92 ± 4.36 | 34.90-45.72

Time from adjustment completion to puncture and fenestration: The mean value was 0.15 ± 0.03 minutes (range, 0.08-0.18 minutes, ie, 5-11 s), indicating stable and highly time-efficient performance of this key step. Total operative time: the mean value was 38.92 ± 4.36 minutes (range, 34.90-45.72 minutes), with controllable duration of the overall workflow. Linear correlation analysis based on the six experimental animals in this study showed no significant linear correlation between fenestration time and total operative time (_R_2 = 0.1154; P = .51), indicating that under the experimental conditions of this study, the duration of this step had no significant impact on the overall operative duration (Fig 5).

Fig 5: Linear correlation analysis between puncture and fenestration time and total operative time. The X-axis represents puncture and fenestration time (minutes), and the Y-axis represents total operative time (minutes); the regression equation is Y = 41.77X + 32.79, with a coefficient of determination R2 = 0.1154 and P = .51, indicating no significant linear correlation between the two (n = 6).

Operator usability feedback rating

As an exploratory assessment to gather preliminary feedback on device design, two attending surgeons rated multiple dimensions of device usability on a 0 to 10-point visual analog scale (0 = unable to operate; 10 = extremely easy to operate) after each procedure. These ratings represent preliminary observational data only and should not be interpreted as a definitive usability validation.

As shown in Table VI, the mean score of all dimensions ranged from 7.17 to 7.83 points. Among them, the dimension of “Directional Accuracy of the Active Steering Directional Puncture Catheter” received the highest score (7.83 ± 0.75 points), followed by “Overall Usability” and “Degree of Procedural Simplification” (both 7.50 points). Verbal feedback from the operators indicated that the integrated design reduced the steps of device exchange, and the active steering function helped to establish a stable puncture angle and boost operational confidence before puncture.

Rating dimension | Surgeon 1 (mean ± standard deviation) | Surgeon 2 (mean ± standard deviation) | Total (mean ± standard deviation)
Overall usability | 6.67 ± 0.58 | 8.33 ± 0.58 | 7.50 ± 1.05
Operability of the cerebral protection device | 6.33 ± 1.15 | 8.00 ± 1.00 | 7.17 ± 1.33
Directional accuracy of the active steering directional puncture catheter | 7.33 ± 0.58 | 8.33 ± 0.58 | 7.83 ± 0.75
Degree of procedural simplification | 7.00 ± 0.00 | 8.00 ± 0.00 | 7.50 ± 0.55
Confidence in safety | 6.33 ± 0.58 | 8.33 ± 0.58 | 7.33 ± 1.21
Overall score (pooled data) | 6.73 ± 0.61 | 8.20 ± 0.56 | 7.47 ± 0.89

These subjective ratings should be interpreted cautiously, as they represent preliminary feedback from a small operator sample and do not constitute a definitive usability validation. They should be considered alongside the objective efficacy indicators reported in this paper.

Safety endpoint evaluation

In this study, the surgical safety of the novel interventional device was systematically evaluated from three dimensions: device integrity, embolus capture efficacy, and vascular histopathological changes. All indicators were assessed via intraoperative monitoring and immediate postoperative examination, with the results presented as follows.

(1) Device integrity

Immediate macroscopic examination was performed on the retrieved integrated system after the procedure. The results showed that the morphological structure of the device remained intact, with no abnormal conditions such as coating exfoliation, filament fracture, or structural damage observed. The integrity and tightness of the key functional components of the device were favorable, with no structural defects that would impair operational efficacy. Thrombus residue was visible inside the embolic protection filter.

(2) Embolus capture analysis

Gentle flushing of the embolic protection filter was performed after the procedure, and a small amount of intravascular thrombus was observed in the flushing fluid. This indicated that the embolic protection filter successfully captured part of the intravascular thrombus debris during the surgical operation, which preliminarily verified its in vivo embolic capture capability (Fig 6).

Fig 6: Macroscopic examination images of the integrity of the retrieved interventional device after the procedure. A, Thrombus debris flushed out from the embolic protection filter after the procedure; B, Captured thrombus residue was visible on the inner surface of the retrieved embolic protection filter, confirming that the system can effectively capture embolic debris generated during the intraoperative operation.

(3) Histopathological evaluation

Immediately after the procedure, iliac artery segments containing the puncture site were harvested. The specimens were fixed with 10% neutral buffered formalin, embedded in paraffin, and sectioned, followed by H&E staining and elastic fiber staining. Microscopic evaluation showed that the continuity of the vascular intima was well preserved, with only occasional focal mild injury, and no obvious laceration or defect was observed (Fig 7, A1-F1). The structure of the vascular elastic lamina was intact, with no manifestations of severe injury such as fracture or displacement. Only mild acute inflammatory cell infiltration was observed around the puncture site, without obvious edema, necrosis, or hematoma formation of the vascular wall. The overall morphological structure of the blood vessels was basically normal, and no severe vascular injury caused by device operation was found (Fig 7, A2-F2).

Fig 7: Results of hematoxylin and eosin (H&E) and elastic fiber staining of iliac artery segments under 4 × 10 magnification. A1-F1, H&E staining of the iliac arteries from the first to sixth experimental pigs, respectively, with mild vascular intimal injury observed in (C1) and (F1); A2-F2, Elastic fiber staining of the iliac arteries from the first to sixth experimental pigs, respectively, with no obvious abnormalities observed in the elastic fiber staining.