An experimental study of the vascular embolism caused by recombinant type III collagen implants and hyaluronic acid

There are significant differences in the reported incidence of vascular complications that result from the injection of different soft tissue fillers. This study aimed to compare the risk of vascular embolism after recombinant type III collagen implants and hyaluronic acid (HA) injection into arteries.

period, and low incidence of complications.Hyaluronic acid (HA) fillers are one of the most popular fillers available. 3[6] Previous study 7 injected 0.1 mL of HA and 0.1 mL of Artecoll (consists of 80% collagen and 20% polymethyl methacrylate) into the central artery of the rabbit ears at a uniform rate within 60 s, and the results revealed that the necrosis areas were significantly increased in the HA groups compared with Artecoll groups.A recombinant type III collagen implant was created as a new product, and no complications related to its injection have been reported.However, it has different physical and chemical properties from Artecoll and its safety needs to be further verified.No relevant controlled experimental studies have compared the risk of vascular complications between HA and collagen.In this study, a rabbit animal model was used to test the risk of vascular embolism and tissue necrosis with HA and collagen.A filler was injected into the central artery of each rabbit ear at a set dose and rate, to simulate the development of vascular complications in clinical practice.The risk comparison between HA and collagen was performed by observing hemodynamic changes, degree of skin necrosis, and histopathological changes.We expected the two groups to significantly differ.

| Research animals and investigational products
In all, 50 healthy white rabbits (2.96 ± 0.21 kg) were used for the study, which required intact rabbit ears with no defects or significant deformities in the ear blood vessels.These rabbits were kept in the Animal Experimentation Center in separate numbered metal cages.
The animals had ad libitum access to food and water, and they were maintained under natural light and suitable temperature and humidity.In the experiment, the combined preparation of xylazine + zoletil was administered by intramuscular injection for anesthesia, and the anesthetic dose was 0.8 mL/kg.Experimental rabbits were observed and photographed on the seventh day after injection and then executed using the CO 2 overdose inhalation method.The rabbits were first anesthetized and then placed in a euthanasia device with an excess of carbon dioxide for execution.The tested products included recombinant type III collagen filler (Jiangsu Trautec Medical Technology Co., Ltd) and HA (Restylane; Q-Med).

| Animal models of intravascular embolism
According to the randomized control method, animals were divided into five groups (C1, C2, C3, C4, and HA), with 20 rabbit ears in each group.The identities of each group were as follows.Group C1: 0.2 mL of collagen was injected at a uniform speed into the central ear artery within 60 s.Group C2: 0.3 mL of collagen was injected at a uniform speed into the central ear artery within 60 s.Group C3: 0.2 mL of collagen was injected at a uniform speed into the central ear artery within 30 s. Group C4: 0.3 mL of collagen was injected at a uniform speed into the central ear artery within 30 s. Group HA: 0.1 mL of HA was injected at a uniform speed into the central ear artery within 60 s.
Anesthesia was carried out using a combination of xylazine + zoletil (1:1) by intramuscular injection, at a dose of 0.8 mL/kg.The dorsal and ventral hair of the rabbit's ears was shaved after anesthesia.The ventral side of the rabbit's ear was illuminated with a bright light so that the main blood vessel course could be clearly distinguished from the dorsal side of the ear.A 27G syringe needle (0.5 mm) was used to inject filler at a uniform speed into the main trunk of the central ear artery, with the injection point located 5 mm distal to the point of bifurcation of the arterial trunk from the lateral branch (Figure 1A, Video S1).All injections were performed by a single plastic surgeon with many years of injection experience and were practiced before the official experiment.An assistant was used to notify the surgeon every 5 s during the injection, to ensure maximum uniformity.

| Observation of intravascular hemodynamic changes and skin necrosis
The following observation time points were set up: before the injection, immediately after the injection, 5 min after the injection, 30 min after the injection, 1 day after the injection, and 7 days after the injection.Changes in hemodynamics in the rabbit ear blood vessels and skin necrosis were carefully observed and photographed using a high-definition camera.The main observations were the vascular recanalization and the degree of skin necrosis, and the evaluation criteria are shown in the following tables.Two senior plastic surgeons performed the evaluation, and a third plastic surgeon was invited to discuss and make a judgment when there was a difference of opinion (Tables 1 and 2).

| Histological examination
After observing, recording, and photographing the rabbits on Day 7 after injection, the proximal, middle, and distal parts of each ear were collected and fixed in 4% formaldehyde.Liquid paraffin was added to the mold, and the tissue samples to be embedded were immersed in paraffin to ensure the correct tissue position.A small amount of liquid paraffin was replenished and then cooled and frozen, to solidify the paraffin for tissue fixation, and then sliced using a microtome at a thickness of about 4-8 um.The tissue sections were placed on slides and soaked in warm water at 40°C to fully stretch the tissue.Dewaxing and hydration were then carried out.The tissue sections were then stained with hematoxylin, differentiated and blued, cleaned, stained with eosin, and dehydrated.The tissue samples were air-dried, sealed, and observed and photographed under a microscope.

| Effect of intra-arterial injection of collagen and HA on vascular recanalization
In all, 15 rabbits in group C1 showed bleeding at the injection site after injection, and 19 rabbits developed hematomas at the injection site within 5 min.However, only one rabbit in group HA showed bleeding at the injection site, and nine rabbits developed hematomas within 5 min (Figure 1B, Video S2).This finding suggests that HA tends to block the blood vessels proximal to the injection site, making it difficult for blood flow to pass through.When collagen enters the blood vessels, the blood vessels proximal to the injection site are not easily blocked completely, and blood still flows through and out of the vessels through the vessel breach.
At 30 min after injection, 17, 16, 17, and 16 rabbits in groups C1, C2, C3, and C4, respectively, had complete recanalization of the central artery, while no rabbits in the HA group had recanalization of the central artery.On Day 1 after injection, all rabbits in the collagen group had complete recanalization of the main trunk of the central ear artery, and no emboli were seen.However, only half of the rabbits in the HA group had complete recanalization of the central ear artery on this day.Embolus formation was seen in the central artery of the non-recanalized rabbit ear.On Day 7 after injection, 1/5 of the rabbits in the HA group still had not recanalized their central ear arteries.In conclusion, when analyzing the vascular recanalization of each group at 5, 30 min, and 1 day after injection, there was a significant difference between the C1 and HA groups (p < 0.05.)However, there was no significant difference between the C1, C2, C3, and C4 groups (p > 0.05).There was also no significant difference in vascular recanalization on Day 7 after injection between the various groups (p > 0.05).
Under the present experimental conditions, the injection dose and injection rate of collagen did not result in significantly different vascular recanalization (Figures 2-4).
Recanalization of the central ear artery in each group of rabbits 5, 30 min, 1, and 7 days after injection are recorded in Figure 5. were completely normal, while 3/20 rabbit ears exhibited mild skin damage, which was recoverable (Figures 1D and 2).There was also a significant difference in skin necrosis between the two groups (p < 0.05).Compared with HA, collagen is less likely to cause severe skin necrosis, and reversible mild skin damage occurs in some cases.
On the other hand, there was no significant difference in the degree of skin damage between the four groups C1, C2, C3, and C4 (p > 0.05).There was also no significant difference in the effect of collagen injection dose and injection rate on skin necrosis under the present experimental conditions.

| Histopathological changes at Day 7 after injection
Local rabbit ear tissues were collected on the seventh day after in- | 2709 were seen at the site of moderate to severe skin injury.Considerable fillers and blood cell aggregation were observed at the injection site (Figure 6).

| DISCUSS ION
Arterial embolism is considered to be the most dangerous and serious complication caused by the application of fillers.The incidence of vascular complications seems to be more frequent than we think.
Results of a global survey of 52 experienced injection staff showed that 62% of them reported one or more vascular complications. 4 Vascular embolism often implies serious consequences, such as skin necrosis and blindness.Unfortunately, effective treatment for vascular embolism is currently very limited.Knowledge of anatomy and injection techniques is important to prevent vascular complications. 8,9The choice of filler is also directly related to the incidence of vascular complications.A meta-analysis of the literature showed that HA induced far more cases of vascular embolism than collagen. 10There are several explanations for this phenomenon: (1) HA particles are larger and have higher hydrophilicity and viscosity, so it is more likely to cause proximal vascular obstruction 11-13 ; (2) some studies have suggested that the two lead to local ischemia by different mechanisms, with collagen possibly leading to inflammatory response and activation of the coagulation system, whereas HA mainly causes mechanical obstruction of the arterial blood supply. 7,14Of course, it needs to be considered that HA has a much higher global usage rate compared to collagen. 15,16So reported complications should be higher than the collagen.
Type I collagen and type III collagen widely exist in skin tissue, especially in dermis. 17For instance, previous researches investigated that embryonic dermis contains about 50% type III collagen, but it reduces to about 15% during post-natal growth. 18,19Type I collagen belongs to the coarse fiber and is usually regarded as structural scaffold to maintain the mechanical strength of skin. 20Type III collagen, which belonging to tiny fiber, provides the elasticity for skin. 21Compared to Type I collagen, Type III collagen injections bring the skin closer to its youthful state and can provide elastic support and contouring.To verify the correlation between the physical properties of the filler and the risk of arterial embolism, the rabbit ear was used as a model in this study.The thin skin and shallow distribution of blood vessels in the rabbit's ear facilitates the observation of vascular embolism.During the injection process, it was observed that significant blood leakage occurred at the injection site within seconds of collagen injection, in both the 0.2 and 0.3 mL groups.
Furthermore, partial recanalization occurred 5 min after injection, and complete recanalization occurred within 30 min.In contrast, after the injection of 0.1 mL of Restylane, the blood vessels were significantly blocked, with no blood leakage and basically no recanalization performance for 30 min.This phenomenon confirms that small particles and low-viscosity collagen enter the blood vessels and easily disperse with the blood, without completely obstructing them.The rate of thrombus formation was significantly higher in the HA group on Day 1 than in the collagen group with a larger injection volume.On Day 7, 95% of the rabbits in the HA group showed moderate to severe skin necrosis, while none of the rabbits in the collagen group showed skin necrosis, and only a few rabbits showed diffuse skin dullness.These results suggest that different fillers have different risks of vascular embolism when they enter the vasculature.Restylane has a higher incidence and severity of vascular complications, while collagen has better safety profile.
The injection dose is closely related to the occurrence of vascular complications.The most severe cases of vascular embolism tend to occur with high-dose filler injections.Expert consensus recommends that injections should be given in small doses at multiple points to ensure continuous needle movement. 11A published study confirmed a higher incidence of vascular complications and skin necrosis in the group injected with 0.2 mL of HA into the central ear artery of rabbits, than in the group injected with 0.1 mL. 7r study found no significant difference in the incidence of vascular complications and skin damage between the 0.2 and 0.3 mL groups when recombinant type III collagen was injected into the central artery of the rabbit ear (p > 0.05).Furthermore, no irreversible skin necrosis occurred in any rabbit ears, of both groups, at 7 days after injection.This result further confirms the safety of collagen implants in soft tissue injections.
No significant difference was found in the incidence of vascular complications and skin damage between the 30s and 60s groups at the same collagen injection volume (p > 0.05).This suggests that the speed of injection may not be a critical factor for the development of vascular embolism.It should be noted that the danger of injecting too quickly is that it leads to an increase in injection pressure, making it easier for the filler to enter the vessel extravascular through a breach in the vessel, thus creating a vascular embolism.
In this experimental model, filler was injected directly in the blood vessel.In addition, the high pressure created by rapid injection encourages the distribution of the filler to more distal vascular branches, making it more likely to undergo retrograde transport

SPSS 16 .
0 software package (IBM Corp.) was used for statistical analysis.The median (quartile) [M (Q1, Q3)] was used for measurement of skewed distributions, and a rate (%) was used for counting data.The measurements of each group were compared and analyzed by independent sample nonparametric tests and a chi square test.The statistical significance level for comparisons was p = 0.05.

F 3 . 2 |
I G U R E 1 (A) Injecting filler uniformly into the vessel with a 27G syringe needle, 5 mm distal to the point of bifurcation between the main trunk and the lateral branch of the central ear artery in rabbits.(B) Hematoma at the injection site within 5 min after injection in the hyaluronic acid (HA) group.(C) Hematoma at the injection site was not seen 5 min after injection in the HA group.(D) Hematoma formation at the injection site after injection in each group.*p < 0.05, n = 20.central arterial trunk was only partially patent at the nodal end, and blood flow was absent in the remaining part Not patent No blood flow was seen in the entire length of the central arterial trunk TA B L E 1 Main observations: vascular recanalization.Effects of intra-arterial collagen and HA injections on skin damage Different skin damage profiles were observed at Day 7 after injection between the groups.Almost all rabbit ears in the HA group exhibited moderate to severe skin necrosis (19/20) (Figure 4), whereas no skin necrosis occurred in the C1 group, with the same injection duration and larger injection dose.In the C1 group, 17/20 rabbit ears jection for HE staining, to observe histopathological changes.The normal rabbit ear area was seen to have intact skin tissue structure, with no inflammation or intravascular foreign bodies.The dull skin areas were found to have intact tissue structure, vascular congestion, inflammatory cell infiltration, and occasional small amounts of intravascular fillers.Inflammation and structureless tissue necrosis TA B L E 2 Main observation markers: skin necrosis.Skin necrosis Description Absent The skin of the rabbit ears was completely normal Mild Localized skin dullness in the rabbit ears Moderate to severe Blackened skin or visible ulcers on the rabbit ears F I G U R E 2 Case 1, Group C1. (A) before injection; (B) immediately after injection; (C) 5 min after injection; (D) 30 min after injection, with most of the central artery trunk recanalized; (E) 1 day after injection, with complete recanalization of the central artery; (F) 7 days after injection, with no signs of skin necrosis.

F I G U R E 3
Case 2, Group C1. (A) before injection; (B) immediately after injection; (C) 5 min after injection; (D) 30 min after injection, complete recanalization of the central arterial trunk; (E) scattered dark patches in the skin 1 day after injection; (F) 7 days after injection, scattered dark patches still existed.

F I G U R E 4
Case 3, hyaluronic acid (HA) group.(A) before injection; (B) immediately after injection; (C) 5 min after injection; (D) 30 min after injection; (E) 1 day after injection, central artery still not recanalized, red emboli visible in blood vessels, and generally dark skin; (F) 7 days after injection, skin ulcers on the dorsal side of the ear; (G) 7 days after injection, skin ulcers on the ventral side of the ear .through the communicating branches.This has important implications in preventing ophthalmic artery embolism after injections in the face.Therefore, in actual clinical practice, slow and uniform injection still has its importance in preventing vascular complications.It is also the standard specification technique recommended by expert consensus. 11

F I G U R E 5
(A) Recanalization of the central ear artery in each group of rabbits 5 min after injection.I: completely recanalized; II: partially recanalized; III: completely non-patent.*p < 0.05, n = 20.(B) Recanalization of the central ear artery in each group of rabbits 30 min after injection.I: completely recanalized; II: partially recanalized; III: completely non-patent.*p < 0.05, n = 20.(C) Recanalization of the central ear artery in each group of rabbits 1 day after injection.I: completely recanalized; II: partially recanalized; III: completely non-patent.*p < 0.05, n = 20.(D) Recanalization of the central ear artery in each group of rabbits 7 days after injection.I: completely recanalized; II: partially recanalized; III: completely non-patent.*p > 0.05, n = 20.F I G U R E 6Column A: normal site of rabbit ear, microscopic skin tissue structure is intact, no inflammation nor intravascular foreign bodies are seen; column B: mild skin damage, dark skin site, microscopic skin tissue structure is intact, vascular congestion, inflammatory cell infiltration, intravascular foreign bodies are occasionally seen; column C: severe skin damage, skin necrosis site, microscopic inflammatory cell infiltration, skin tissue showing no structural necrosis; column d: specimens with bleeding at the injection site with a large amount of extravascular filling and blood cell aggregation.