**3. Selection of an experimental saccular aneurysm model**

#### **3.1. Concept of experimental saccular aneurysm**

48 Aneurysm

ophthalmic artery. The main branch of the internal maxillary artery is the middle meningeal artery. The intracranial internal carotid artery (ICA) divides into the ophthalmic arteries, cranial, and caudal branches. The cranial branch runs forward towards the uncus, where it divides into the anterior choroidal artery and middle cerebral artery (MCA) trunk, and then continues up to the chiasm, where it unites with the contralateral cranial branch to form a common anterior cerebral artery trunk that separates again at the level of the corpus callosum. The common anterior trunk originates from the lateral artery of the olfactory bulb, which leads to the ethmoidal branches of the cribiform plate. The MCA runs along the lateral cerebral sulcus and divides into the posterior ophthalmic artery, large posterior branch, and large anterior and middle branches, in addition to the small olfactory bulb branches. The caudal branch of the ICA supplies most of the blood flow of the basilar artery (BA) and leads to the following branches: posterior communicating artery, small medial geniculate body branches, large anterior quadrigeminal body, small branches of the posterior side of the uncus, and the posterior segment of the corpus callosum. The cerebellar artery might originate from the ending of the ICA or the BA and connects to several branches of the brainstem. The BA is formed by the fusion of the arteries of the first spinal nerves and divides (on the ventral surface of the trapezoid body) into two vessels that reunite at the upper margin of the pons. In addition, the BA gives small lateral branches, the cerebellar artery and the perforating branches. The arteries of the first spinal nerve then

reunite at a lower level and form the ventral spinal artery[49].

**1 2**

**3 4**

**Figure 1.** Graphic representation of the visceral vascularization of the neck of rabbits. 1- superior laryngeal artery, 2- superior branches, 3- superior thyroid artery, 4- cricothyroid branch, 5- bronchial

Regarding the system of intracranial anastomoses in rabbits, the collateral circulation is very different from that of dogs. The internal maxillary artery originates from the orbital branches, which end at the ophthalmic branch and represents an insufficient anastomotic pathway. The anastomotic branches between the orbital and internal carotid arteries are too

branch and 6- tracheoesophageal branch. Modified from Bugge, 1967[2].

**5 <sup>6</sup> <sup>6</sup>** Experimental saccular aneurysms are induced aneurysms intended to reproduce the histological, geometric, and hemodynamic characteristics of human intracranial aneurysms.

#### **3.2. Characteristics of an ideal model of experimental saccular aneurysm**

With the rise of endovascular treatment of human intracranial aneurysms – by means of embolization using platinum microcoils[16] – experimental models of saccular aneurysm are encouraged to adapt to this novel therapeutic modality by meeting the following criteria: 1) demonstration of long-term permeability in untreated control species, 2) development in animal species with a coagulation system similar to that of humans, 3) simulation of the morphology of arterial bifurcation, terminal artery, or other aneurysmal types that expose the aneurysm neck to high hemodynamic tension, 4) development in vessels with a similar size to human intracranial vessels, 5) development without the need of local surgery to minimize the repair/wound healing response, which might confound the results of the experiment with the natural increase of the biological activity characteristic of several

embolization materials such as: coils, fluid agents, etc., and 6) simulation of the limitations met by embolization of human aneurysms using such materials[39].

## **3.3. Main models of experimental saccular aneurysm**

German and Black (1954) were the first researchers to produce experimental aneurysms using a surgical construction of saccular aneurysms on the common carotid artery of dogs. Such aneurysms mimicked the ones occurring on the lateral wall and were frequently used in hemodynamic studies; however, they produced fibrosis at the suture site, which was a disadvantage[20]. Since then, surgical models have evolved with the culmination of the swine model (1994), consisting of a graft of the venous pouch onto the common carotid artery (CCA) of pigs. This method produces lateral wall aneurysms, but includes disadvantages such as venous histology, induction of intense fibrosis at the suture site, and low hemodynamic tension[16].

In addition to the surgical method, chemical induction might also be used in the construction of saccular aneurysms. The main proponent of this technique was Hashimoto (1970), who induced arterial wall weakening in rats by ingestion of 3-beta-aminopropionitrile, a toxic agent extracted from the seeds of the sweat pea (*Latyrus odoratus*), which destroys the elastic fibers and collagen of the arteries of rats[19]. In addition, Hashimoto ligated one of the common carotid arteries and induced arterial hypertension in rats (via nephrectomy, intake of saline solution, and high doses of corticosteroids) to cause greater hemodynamic tension on the weakened arterial wall[30]. This technique was the first to produce successful intracranial saccular aneurysms at the bifurcations of the cerebral arterial circle. Nonetheless, the aneurysms were too small and were not useful for the development of surgical techniques nor for the study of intra-aneurysmal hemodynamic alterations[26, 29].

#### *3.3.1. Surgical models*

The technique used in the surgical construction of experimental aneurysm is based on grafting a venous pouch (usually taken from the external jugular vein) onto the common carotid artery. The main advantage of this approach is that the constructed aneurysms exhibit hemodynamic features that are very similar to those of humans. the disadvantages of constructed aneurysms include their venous histology and resistance to rupture.

With regard to the construction site, the graft might be placed on the lateral wall or at bifurcations. There are five main techniques to construct lateral wall aneurysms:


The main model for the construction of bifurcation aneurysms was performed using Forrest and O´Rielly's technique, in which the left common carotid artery of rabbits was partially anastomosed with the right common carotid artery. Next, a venous pouch (taken from the external jugular, anterior facial, or posterior facial vein) was grafted onto the knot formed by the union of the arterial anastomoses. The advantage of this technique was that unlike the lateral wall aneurysms, it did not induce aneurysmal thrombosis (**figure 3**)[24].

**Figure 3.** Graphic representation of the main surgical models of experimental saccular aneurysm. (a) Lateral wall, (b) bifurcation (RCCA – right common carotid artery, EJV – external jugular vein, LCCA – left common carotid artery).

#### *3.3.2. Other experimental models of aneurysms*

In addition to the abovementioned techniques, other methods have been attempted to construct saccular aneurysms, such as hyper-flow (through the creation of arteriovenous fistulas), trauma (traumatic puncture of the arterial wall or using CO2 laser), and chemical wall injury (by injecting nitrogen mustard or other substances directly inside the arterial wall) [51]. All of these techniques are less efficient than chemical induction and surgical construction. Despite these attempts at the construction of an experimental model of saccular aneurysm, none of these methods was able to reproduce all of the histopathological, geometric, and hemodynamic features of human intracranial saccular aneurysms [51-54]. Nevertheless, the enzymatic method has stood out in recent years.

#### *3.3.3. Enzymatic models*

50 Aneurysm

embolization materials such as: coils, fluid agents, etc., and 6) simulation of the limitations

German and Black (1954) were the first researchers to produce experimental aneurysms using a surgical construction of saccular aneurysms on the common carotid artery of dogs. Such aneurysms mimicked the ones occurring on the lateral wall and were frequently used in hemodynamic studies; however, they produced fibrosis at the suture site, which was a disadvantage[20]. Since then, surgical models have evolved with the culmination of the swine model (1994), consisting of a graft of the venous pouch onto the common carotid artery (CCA) of pigs. This method produces lateral wall aneurysms, but includes disadvantages such as venous histology, induction of intense fibrosis at the suture site, and

In addition to the surgical method, chemical induction might also be used in the construction of saccular aneurysms. The main proponent of this technique was Hashimoto (1970), who induced arterial wall weakening in rats by ingestion of 3-beta-aminopropionitrile, a toxic agent extracted from the seeds of the sweat pea (*Latyrus odoratus*), which destroys the elastic fibers and collagen of the arteries of rats[19]. In addition, Hashimoto ligated one of the common carotid arteries and induced arterial hypertension in rats (via nephrectomy, intake of saline solution, and high doses of corticosteroids) to cause greater hemodynamic tension on the weakened arterial wall[30]. This technique was the first to produce successful intracranial saccular aneurysms at the bifurcations of the cerebral arterial circle. Nonetheless, the aneurysms were too small and were not useful for the development of surgical techniques

The technique used in the surgical construction of experimental aneurysm is based on grafting a venous pouch (usually taken from the external jugular vein) onto the common carotid artery. The main advantage of this approach is that the constructed aneurysms exhibit hemodynamic features that are very similar to those of humans. the disadvantages of

With regard to the construction site, the graft might be placed on the lateral wall or at

2. Non-ligated venous pouch with side-to-side anastomosis to the artery (variation of the

5. End-to-side anastomosis of the venous pouch. The main advantage of this technique is the short-lasting clamping of the common carotid artery that thus avoids endothelial

constructed aneurysms include their venous histology and resistance to rupture.

bifurcations. There are five main techniques to construct lateral wall aneurysms:

3. End-to-side anastomosis of the vein onto the artery with ligated venous pouch. 4. Side-to-side anastomosis of the vein onto the artery with ligated venous pouch.

1. Non-ligated venous pouch with end-to-side anastomosis to the artery.

met by embolization of human aneurysms using such materials[39].

nor for the study of intra-aneurysmal hemodynamic alterations[26, 29].

**3.3. Main models of experimental saccular aneurysm** 

low hemodynamic tension[16].

*3.3.1. Surgical models* 

former).

damage and vasospasm[21].

#### *3.3.3.1. Elastase-induced model*

#### *3.3.3.1.1. Mechanisms of action of elastase in aneurysm formation*

The formation of saccular aneurysms depends on several mechanisms, including inflammatory reaction, weakening of the arterial wall, and hemodynamic tension. Enzymatic imbalance and inflammatory activity are some of the potential causes involved in aneurysm formation in humans. Anidjar (1992) perfused the abdominal aorta of a group of Wistar rats with pancreatic elastase from swine and used thioglycollate plus plasmin (activators of the inflammatory response) in another group of animals. Both groups exhibited an inflammatory reaction, elastic lamina fragmentation, and formation of fusiform aneurysms similar to those

that occur in humans. The inflammatory activity was stronger in the elastase group (achieving its peak on the sixth day) and produced macrophages, polymorphonuclear cells, helper and suppressor T lymphocytes in the arterial wall. Combined with the progression of the inflammatory activity, the diameter of the abdominal aorta increased[55]. Halpern (1994) established the sequence and synchrony of induction of the inflammatory response. Elastase induces injury of the arterial wall, which triggers an initial inflammatory response. The inflammatory cells then activate endogenous proteinases (molecular weight between 50 and 90 kD) and the destruction of elastin and collagen, in addition to aortic dilation. Halpern's study showed that the rupture of elastin and its contact with macrophages are the main events in the activation of endogenous proteinases, which results in increased tissue destruction [56].

Although inflammatory activity might lead to destruction of the elastic fibers and a weakening of the arterial wall, its role in the development of saccular aneurysms has not been fully established. Other mechanisms may also participate in aneurysm formation such as alterations of the mechanical properties of arteries together with the hemodynamic tension on the vascular wall, which can produce aneurysms by themselves. Miskolczi (1997) demonstrated this phenomenon in an in vitro study, in which the common carotid arteries of swine and sheep were isolated and their walls were digested using pancreatic elastase from swine. Next, the arterial segments were placed between a pulsatile flow artificial pump and a series of test tubes, which allowed the control of variables such as flow, pulsation, and pressure without inducing the inflammatory response that occurs in in vivo studies. Consequently, small saccular aneurysms appeared at the sites where the elastin was damaged and hemodynamic tension was exerted on the weakened arterial wall[57].

#### *3.3.3.1.2. Creation and improvement of the elastase-induced model*

Based on studies of experimental aneurysm creation using elastase [55, 56], Cawley et al. (1996) developed a new experimental model of lateral wall aneurysms in rabbits. This model consisted of dissecting the neck of rabbits, ligating the proximal segment of the external carotid artery, and performing intra-arterial perfusion of pancreatic elastase from swine. Three weeks later, saccular aneurysms were formed, which, from angiographic and histological perspectives, were very similar to those in humans. However, the lumen remained patent in only 40% of the aneurysms, because lateral aneurysms do not originate from the type of hemodynamic stress and intra-aneurysmal blood flow that occur at the bifurcations of the human cerebral arteries[58].

Cloft et al. (1999) improved this model by producing greater hemodynamic stress on the left common carotid artery (LCCA), which was directly hit by the blood flow from the ascending aorta in two-thirds of the rabbits. This technique is fully endovascular and consists of insufflating a balloon at the origin of the LCCA and isolating a small arterial segment for intraluminal infusion of bovine pancreatic elastase for 30 minutes. Angiographic control was performed by the dissection and retrograde puncture of the femoral arteries. This method succeeded in producing aneurysms with an average size of 3.0 mm x 5.0 mm whose lumen remained patent up to three months after creation. From a microscopic point of view, all of the aneurysms exhibited intact endothelium, the absence of an inflammatory response, moderately damaged elastic lamina inside of the aneurysm (but undamaged at the neck), and apical thrombus. No animal exhibited neurological sequelae (due to the intracranial collateral vessels network) or showed systemic signs of elastase intoxication[59].

52 Aneurysm

that occur in humans. The inflammatory activity was stronger in the elastase group (achieving its peak on the sixth day) and produced macrophages, polymorphonuclear cells, helper and suppressor T lymphocytes in the arterial wall. Combined with the progression of the inflammatory activity, the diameter of the abdominal aorta increased[55]. Halpern (1994) established the sequence and synchrony of induction of the inflammatory response. Elastase induces injury of the arterial wall, which triggers an initial inflammatory response. The inflammatory cells then activate endogenous proteinases (molecular weight between 50 and 90 kD) and the destruction of elastin and collagen, in addition to aortic dilation. Halpern's study showed that the rupture of elastin and its contact with macrophages are the main events in the

activation of endogenous proteinases, which results in increased tissue destruction [56].

damaged and hemodynamic tension was exerted on the weakened arterial wall[57].

Based on studies of experimental aneurysm creation using elastase [55, 56], Cawley et al. (1996) developed a new experimental model of lateral wall aneurysms in rabbits. This model consisted of dissecting the neck of rabbits, ligating the proximal segment of the external carotid artery, and performing intra-arterial perfusion of pancreatic elastase from swine. Three weeks later, saccular aneurysms were formed, which, from angiographic and histological perspectives, were very similar to those in humans. However, the lumen remained patent in only 40% of the aneurysms, because lateral aneurysms do not originate from the type of hemodynamic stress and intra-aneurysmal blood flow that occur at the

Cloft et al. (1999) improved this model by producing greater hemodynamic stress on the left common carotid artery (LCCA), which was directly hit by the blood flow from the ascending aorta in two-thirds of the rabbits. This technique is fully endovascular and consists of insufflating a balloon at the origin of the LCCA and isolating a small arterial segment for intraluminal infusion of bovine pancreatic elastase for 30 minutes. Angiographic control was performed by the dissection and retrograde puncture of the femoral arteries. This method succeeded in producing aneurysms with an average size of 3.0 mm x 5.0 mm whose lumen remained patent up to three months after creation. From a microscopic point of view, all of the aneurysms exhibited intact endothelium, the absence of an inflammatory response, moderately damaged elastic lamina inside of the aneurysm (but undamaged at the neck), and

*3.3.3.1.2. Creation and improvement of the elastase-induced model* 

bifurcations of the human cerebral arteries[58].

Although inflammatory activity might lead to destruction of the elastic fibers and a weakening of the arterial wall, its role in the development of saccular aneurysms has not been fully established. Other mechanisms may also participate in aneurysm formation such as alterations of the mechanical properties of arteries together with the hemodynamic tension on the vascular wall, which can produce aneurysms by themselves. Miskolczi (1997) demonstrated this phenomenon in an in vitro study, in which the common carotid arteries of swine and sheep were isolated and their walls were digested using pancreatic elastase from swine. Next, the arterial segments were placed between a pulsatile flow artificial pump and a series of test tubes, which allowed the control of variables such as flow, pulsation, and pressure without inducing the inflammatory response that occurs in in vivo studies. Consequently, small saccular aneurysms appeared at the sites where the elastin was Kallmes et al. (1999) modified this method by creating additional hemodynamic tension on the proximal segment of the right common carotid artery (RCCA), which is located between the brachiocephalic artery and ascending aorta, and mimics a "bifurcation aneurysm." In addition, the long curvature of the brachiocephalic artery increased the hemodynamic tension at the origin of the RCCA compared to the LCCA. Further modification of this model consisted of reducing the time of enzymatic digestion to 20 minutes (**figure 4**).

**Figure 4.** Graphic representation of the endovascular elastase-induced aneurysm construction technique. AoA – aortic arch, RCCA – right common carotid artery and LCCA – left common carotid artery. Modified from Hoh, 2004[60].

These technical modifications resulted in experimental aneurysms similar to those observed in humans with regard to the arterial origin, shape, hemodynamics, and patency. The high hemodynamic tension caused by the long curvature of the brachiocephalic artery makes these experimental aneurysms similar to those occurring in the ophthalmic segment of the human internal carotid artery[40]. Altes et al. (2000) used the RCCA for the intraluminal infusion of pancreatic elastase from swine in rabbits and obtained aneurysms in 89% of the animals. Two weeks later, the elastic lamina ruptured and aneurysms were formed (average dimensions of 4.5 mm x 7.5 mm), with organized thrombus in the aneurysm dome, whereas the elastic lamina was undamaged in the walls of the parent arteries. The cells present in the organized thrombus exhibited features of smooth muscle cells and fibroblasts. Ten weeks later, no significant alterations were observed. The execution of this technique required less than one hour, and although it included surgical procedures (e.g., section of the RCCA), this technique exhibited lower morbidity and mortality compared to the use of the LCCA[43].

From a technical perspective, it is noteworthy that the concentration of elastase and the time of incubation exert a partial effect on the size of the aneurysms. One study compared animals that were not subjected to elastase to animals that were subjected to low, medium, and high concentrations of this drug, under variable durations. The rabbits that were not subjected to elastase exhibited complete thrombosis of the arterial stump and did not form

aneurysms, whereas the rabbits that were given elastase in progressive concentrations formed aneurysms. The increase of the elastase dose above a given value did not influence the size of the aneurysms; however, high concentrations of elastase induced the dilation of the parent artery and resulted in a more complex geometry of the aneurysm neck, which is closer to that observed in human aneurysms. Low concentration (25%) of elastase induced aneurysms without dilation of the adjacent artery[61].

Hoh et al. (2004) developed a simpler technique of construction and obtained aneurysms similar to those previously mentioned. The first simplification consisted of the use of a 24 gauge angiocatheter (instead of an introducer) and transitory occlusion of the origin of the RCCA using a neurosurgical clamp (instead of a balloon)[60]. The second simplification was achieved using an accurate neurological assessment of the rabbits using a four-point scale to rate the observed movements of the rabbits on a flat surface to verify whether paresis of the legs or abnormal gait occurred (movements in a circle or difficulty to walk). Accordingly, the animals were rated as grade I – no neurological deficit; grade II: minimal of suspected neurological deficit; grade III: mild neurological deficit without abnormal motion; and grade IV: remarkable neurological deficit and abnormal motion[62].

Although the studies performed to date have not reported any loss of animals, Möller-Hartmann et al. (2003) found a mortality of 25% due to the accidental passage of elastase into the superior thyroid artery with an aberrant origin or into the tracheoesophageal branch, which originated in the common carotid artery, resulting in hemorrhagic necrosis of the trachea[63]. Another source of undesirable distribution of elastase and tracheal necrosis is the anomalous origin of the tracheobronchial artery, which can be identified in angiographies as a small branch perpendicular to the proximal part of the RCCA, and runs medially towards the trachea[64]. Therefore, the elimination of those anomalous vessels (by ligation, coagulation or placing of the introducer lower inside the RCCA) is crucial for success in aneurysm creation by intraluminal infusion of elastase[63].

In addition to the problem posed by aberrant vessels, Krings et al. (2003) identified two additional potential causes of failure of the elastase model. The first potential cause depends on how elastase is injected through the introducer. Thus, instead of elastase, the blood column of the introducer dead space is pushed into the arterial lumen. Furthermore, the authors observed that doses of 100U of elastase were usually lethal. To address these problems, the authors reduced the dose of elastase to 20 U and performed a contrast injection test to detect aberrant arteries as follows: after occluding with a balloon in the proximal area of the RCCA, a non-ionic contrast material was injected (by means of an introducer) inside the RCCA, and the contrast column was verified for two minutes. If the contrast material remained, without washing out or dilution for two minutes, the test was deemed to be negative, i.e., there were no anomalous vessels. Otherwise, the test was deemed to be positive, and the introducer was advanced to a more proximal site of the RCCA where the contrast washing out or dilution no longer occurred. When these procedures were applied, none of the animals died, and all developed aneurysms. The full duration of this procedure was 40 minutes. The problem posed by the blood column and contrast material inside of the introducer was resolved by performing continuous suction using a syringe[65].

Prospective studies on the morphology and viability of elastase-induced aneurysms in rabbits require serial high-quality angiographic control. Three routes are currently used: the femoral arteries, left external auricular vein, and left central auricular artery. Miskolczi et al. (2005) suggested performing a retrograde puncture of the left central auricular artery as the best route, because the femoral artery is narrow and fragile and thus exhibits a high risk of injury and definitive loss. In addition, retrograde femoral catheterization requires the dissection of the groin, arteriotomy, and subsequent ligation with permanent vascular occlusion, thus making subsequent angiography at this site impossible. Puncture of the left external auricular vein allows for repeated injections of contrast material, but the resulting images exhibit low spatial resolution and frequent motion-related artifacts. In contrast, the left central auricular artery allows for repeated injections, high-quality images, and excellent visualization of the brachiocephalic trunk vessels because rabbits usually exhibit LCCA of bovine origin; thus, when the contrast material is injected into the left central auricular artery, the brachiocephalic trunk and its branches immediately become filled. When the LCCA originates directly from the aortic arch or from a common origin with the brachiocephalic trunk, but the angle is unfavorable, the contrast material only fills the distal aortic arch. The anatomy of approximately 70% - 80% of white New Zealand rabbits is favorable for retrograde injection in the left central auricular artery; therefore, pre-selection is important to exclude animals with unfavorable anatomy from studies[66].

#### *3.3.3.1.3. Morphological and geometric features*

54 Aneurysm

aneurysms, whereas the rabbits that were given elastase in progressive concentrations formed aneurysms. The increase of the elastase dose above a given value did not influence the size of the aneurysms; however, high concentrations of elastase induced the dilation of the parent artery and resulted in a more complex geometry of the aneurysm neck, which is closer to that observed in human aneurysms. Low concentration (25%) of elastase induced

Hoh et al. (2004) developed a simpler technique of construction and obtained aneurysms similar to those previously mentioned. The first simplification consisted of the use of a 24 gauge angiocatheter (instead of an introducer) and transitory occlusion of the origin of the RCCA using a neurosurgical clamp (instead of a balloon)[60]. The second simplification was achieved using an accurate neurological assessment of the rabbits using a four-point scale to rate the observed movements of the rabbits on a flat surface to verify whether paresis of the legs or abnormal gait occurred (movements in a circle or difficulty to walk). Accordingly, the animals were rated as grade I – no neurological deficit; grade II: minimal of suspected neurological deficit; grade III: mild neurological deficit without abnormal motion; and grade

Although the studies performed to date have not reported any loss of animals, Möller-Hartmann et al. (2003) found a mortality of 25% due to the accidental passage of elastase into the superior thyroid artery with an aberrant origin or into the tracheoesophageal branch, which originated in the common carotid artery, resulting in hemorrhagic necrosis of the trachea[63]. Another source of undesirable distribution of elastase and tracheal necrosis is the anomalous origin of the tracheobronchial artery, which can be identified in angiographies as a small branch perpendicular to the proximal part of the RCCA, and runs medially towards the trachea[64]. Therefore, the elimination of those anomalous vessels (by ligation, coagulation or placing of the introducer lower inside the RCCA) is crucial for

In addition to the problem posed by aberrant vessels, Krings et al. (2003) identified two additional potential causes of failure of the elastase model. The first potential cause depends on how elastase is injected through the introducer. Thus, instead of elastase, the blood column of the introducer dead space is pushed into the arterial lumen. Furthermore, the authors observed that doses of 100U of elastase were usually lethal. To address these problems, the authors reduced the dose of elastase to 20 U and performed a contrast injection test to detect aberrant arteries as follows: after occluding with a balloon in the proximal area of the RCCA, a non-ionic contrast material was injected (by means of an introducer) inside the RCCA, and the contrast column was verified for two minutes. If the contrast material remained, without washing out or dilution for two minutes, the test was deemed to be negative, i.e., there were no anomalous vessels. Otherwise, the test was deemed to be positive, and the introducer was advanced to a more proximal site of the RCCA where the contrast washing out or dilution no longer occurred. When these procedures were applied, none of the animals died, and all developed aneurysms. The full duration of this procedure was 40 minutes. The problem posed by the blood column and contrast material inside of the

aneurysms without dilation of the adjacent artery[61].

IV: remarkable neurological deficit and abnormal motion[62].

success in aneurysm creation by intraluminal infusion of elastase[63].

introducer was resolved by performing continuous suction using a syringe[65].

The elastase model efficiently reproduces aneurysms similar to ones that occur in the ophthalmic segment of the human internal carotid artery with regard to width, height, neck size, and diameter of the parent artery. These characteristics were very well established by Short et al. (2001), who prospectively studied 40 rabbits and observed that the size of the aneurysmal cavities afforded by the elastase model was appropriate for preclinical tests of endovascular occlusion techniques and devices. The authors measured the width (points in the cavity exhibiting the maximal width), height (measurement of the aneurysmal dome to the mid-portion of a line connecting the proximal and distal portions of the aneurysm neck), neck (maximal diameter between the proximal and distal portions of the aneurysm orifice), the diameter of the parent artery (diameter of the artery just proximal to the aneurysm neck), and the dome/neck ratio (maximal dome width/neck width). In addition, they classified the aneurysms as small (2.0 mm – 4.9 mm), medium-sized (5.0 mm – 9.9 mm), or large (10.0 – 16.0 mm). Moreover, the neck was classified as small (< 4 mm) or wide (> 4 mm). Two weeks later, all of the animals had survived, none showed clinical evidences of neurological insult, and exhibited aneurysms at the apex of the long curve of the brachiocephalic artery, with an elongated shape, and a height greater than the width. Medium-sized (50%) and large (42.5%) aneurysms with small necks (55%) prevailed. The average width of the cavity was 4.1 ± 1.2 mm, which varied between 2.5 and 7.1 mm, and the average height was 8.8 ± 2.6 mm, which varied between 3.0 and 15.6 mm. A dome/neck ratio > 1 was observed in 50% of the aneurysms with an average value of 1.13 ± 0.5, and the average diameter of the parent artery was 4.3 ± 1.4 mm. Although these measures were similar to those of human aneurysms, they did not reproduce all of the corresponding morphological characteristics, which are difficult to quantify for many reasons[44].

Short-term follow-up of elastase-induced aneurysms showed that their dimensions increased gradually up to the end of the first month after creation and then become stable. The average measurements of the dome width and length at days 3 and 28 after induction were (3.2 ± 0.6 mm; 5.0 ± 0.9 mm) and (6.0 ± 1.3 mm; 10.0 ± 2.2 mm), respectively. Conversely, the aneurysms that were not incubated with elastase progressively retracted and formed thrombi inside. Because a millimeter-scale was used and the differences found were small, the authors considered the low resolution of intravascular angiography, radiographic projections used, and variations of the cardiac cycle that promoted different intra-aneurysmal pressures to be potential sources of variation and the lack of histological correlation to be a limitation of the study[67]. Ding et al. (2006) studied the long-term permeability of elastase-induced aneurysms and observed that the aneurysmal cavity remained patent and without thrombi for up to two years after creation and that after the first month, their dimensions (width, height, and neck width) did not exhibit significant variation [68].

The size of the neck has paramount importance when testing endovascular devices, as well as in the study of the physiopathology of aneurysms, and might be modified during the construction of experimental aneurysms. This finding was revealed by Ding et al. (2005), who observed that the size of the neck might be controlled by adjusting the position of the balloon during incubation with elastase. When the balloon is placed high, that is, half inside the proximal RCCA and half inside the subclavian and brachiocephalic arteries, the neck of the resulting aneurysms is narrow (< 4 mm). When the balloon is placed low, that is, exclusively inside of the subclavian and brachiocephalic arteries, the neck of the resulting aneurysms is wide (> 4 mm). The authors further observed that the position of the balloon did not influence the length of the aneurysms and that the balloons that were placed low did not always result in wide necks[69].

In addition to the low position of the balloon, the geometric relationship between the longest axis of the aneurysms and the axis of the parent artery played an important role in the determination of local hemodynamics and the final architecture of aneurysms. Onizuka et al. (2006) compared the angle formed by the longest axis of aneurysms and the axis of the parent artery immediately and three months after aneurysm construction. The authors found a positive correlation between the neck size and the dome height. In addition, the dome height was proportional to the angle formed by the brachiocephalic artery and the aneurysm neck. Therefore, the authors concluded that the larger the angle, the greater the hemodynamic stress caused by the blood flow on the distal neck and the aneurysm bottom[70].

The volume of elastase-induced aneurysms might also be adjusted by the position of the RCCA ligation so that high ligations might create relatively larger aneurysms compared to the ones produced by low ligations. Ding et al. (2007) prospectively studied the influence of the height of the RCCA ligation on the volume of aneurysms. Ligations were rated lower when the height of the ligation point was 10-mm away from the origin of the RCCA and high when the ligation point was 15-mm away from the origin of the RCCA. The same authors applied the formula for the volume of cylinders to calculate the volume of aneurysms because the shape of the created aneurysms was cylindrical. The aneurysms with higher ligations exhibited larger volumes (102.4 ± 54.8 mm3) compared to the ones with lower ligations (36.6 ± 26.8 mm3). In addition, the aneurysms with higher ligations exhibited larger dimensions such as the neck (3.3 ± 0.8 mm), width (3.7 ± 0.7 mm), and height (9.0 ± 1.7 mm). The authors attributed these results to a larger cavity space of aneurysms with higher ligation, in addition to probable greater hemodynamic stress on the aneurysms. Finally, according to those authors, no animals died due to the accidental passage of elastase (through aberrant vessels) in the case of aneurysms with higher ligation[69].

#### *3.3.3.1.4. Histology*

56 Aneurysm

Short-term follow-up of elastase-induced aneurysms showed that their dimensions increased gradually up to the end of the first month after creation and then become stable. The average measurements of the dome width and length at days 3 and 28 after induction were (3.2 ± 0.6 mm; 5.0 ± 0.9 mm) and (6.0 ± 1.3 mm; 10.0 ± 2.2 mm), respectively. Conversely, the aneurysms that were not incubated with elastase progressively retracted and formed thrombi inside. Because a millimeter-scale was used and the differences found were small, the authors considered the low resolution of intravascular angiography, radiographic projections used, and variations of the cardiac cycle that promoted different intra-aneurysmal pressures to be potential sources of variation and the lack of histological correlation to be a limitation of the study[67]. Ding et al. (2006) studied the long-term permeability of elastase-induced aneurysms and observed that the aneurysmal cavity remained patent and without thrombi for up to two years after creation and that after the first month, their dimensions (width,

The size of the neck has paramount importance when testing endovascular devices, as well as in the study of the physiopathology of aneurysms, and might be modified during the construction of experimental aneurysms. This finding was revealed by Ding et al. (2005), who observed that the size of the neck might be controlled by adjusting the position of the balloon during incubation with elastase. When the balloon is placed high, that is, half inside the proximal RCCA and half inside the subclavian and brachiocephalic arteries, the neck of the resulting aneurysms is narrow (< 4 mm). When the balloon is placed low, that is, exclusively inside of the subclavian and brachiocephalic arteries, the neck of the resulting aneurysms is wide (> 4 mm). The authors further observed that the position of the balloon did not influence the length of the aneurysms and that the balloons that were placed low did

In addition to the low position of the balloon, the geometric relationship between the longest axis of the aneurysms and the axis of the parent artery played an important role in the determination of local hemodynamics and the final architecture of aneurysms. Onizuka et al. (2006) compared the angle formed by the longest axis of aneurysms and the axis of the parent artery immediately and three months after aneurysm construction. The authors found a positive correlation between the neck size and the dome height. In addition, the dome height was proportional to the angle formed by the brachiocephalic artery and the aneurysm neck. Therefore, the authors concluded that the larger the angle, the greater the hemodynamic

The volume of elastase-induced aneurysms might also be adjusted by the position of the RCCA ligation so that high ligations might create relatively larger aneurysms compared to the ones produced by low ligations. Ding et al. (2007) prospectively studied the influence of the height of the RCCA ligation on the volume of aneurysms. Ligations were rated lower when the height of the ligation point was 10-mm away from the origin of the RCCA and high when the ligation point was 15-mm away from the origin of the RCCA. The same authors applied the formula for the volume of cylinders to calculate the volume of aneurysms because the shape of the created aneurysms was cylindrical. The aneurysms with higher ligations exhibited larger volumes (102.4 ± 54.8 mm3) compared to the ones with lower ligations (36.6 ± 26.8 mm3). In addition, the aneurysms with higher ligations exhibited

stress caused by the blood flow on the distal neck and the aneurysm bottom[70].

height, and neck width) did not exhibit significant variation [68].

not always result in wide necks[69].

Abruzzo et al. (1998) compared the histological characteristics between lateral wall aneurysms (produced by means of elastase incubation in the external carotid artery of rabbits) and lateral wall aneurysms constructed by grafting a venous pouch onto the common carotid artery of pigs. Both experimental aneurysms were compared to human aneurysms with 5 – 10 mm of diameter (recently ruptured and obtained at autopsy), whose main characteristics included: 1) a complete absence of the internal elastic lamina in the aneurysms, and abrupt termination of the internal elastic lamina of the parent artery at the margins of the saccular orifice; 2) complete absence of the tunica media in the aneurysms and abrupt termination of the tunica media of the parent artery at the margins of the aneurysmal orifice; 3) absence of intramural inflammatory reaction in the aneurysms; 4) absence of neointimal fibromuscular proliferation; 5) a sac wall thickness of 51 μm and a neck thickness of 52 μm. In three out of the five studied aneurysms, one-third of the aneurysmal cavity was filled by a thrombus at different stages of organization and firmly adhered to the point of rupture. The elastase-induced aneurysms exhibited an abrupt termination of the internal elastic lamina at the margins of the saccular orifice, but the tunica medica was undamaged and continued into the interior of the saccular part of the aneurysms. The sac walls exhibited a mild to moderately inflammatory cellular (monocytes and neutrophils) response and a mild fibromuscular response. The thickness of the neck was 49 μm, and the thickness at the sac wall was 44 μm. An unorganized thrombus filled one-third of the aneurysmal cavity in two out of the four investigated rabbits. The aneurysms constructed using a venous pouch exhibited a well-developed elastic lamina, and the tunica media extended into the sac wall. The wall of the venous pouch contained remarkable inflammatory infiltrate (monocytes and neutrophils) and extreme degrees of fibromuscular proliferation completely across the aneurysm wall, resulting in a remarkable neointimal thickening and luminal narrowing. The thickness of the dome wall was 228 μm, and the thickness at the neck was 350 μm. Thus, the authors concluded that from a histological perspective, the elastase-induced aneurysms were the ones most similar to human aneurysms, in addition to exhibiting little spontaneous fibromuscular response compared to the surgical model with venous pouch grafting[71]. Accordingly, the elastase-induced model is currently used in tests for endovascular devices[39-42].

#### *3.3.3.2. Papain-induced model*

Although the damage of elastic fibers induced by swine pancreatic elastase resulted in experimental aneurysms similar to those appearing in the ophthalmic segment of the human internal carotid artery, they are small (<5 mm), which is not completely consistent with the actual clinical characteristics of human aneurysms, where the aneurysms are larger than 5 mm. To overcome this limitation, Chinese researchers tested an association between elastase and collagenase in the *in vitro* pre-digestion of an arterial pouch grafted onto the aortic arch

of rabbits; however, that model exhibited a higher tendency to spontaneously rupture[72]. To produce saccular aneurysms larger than 5 mm, De Oliveira et al. (2011) infused the papain enzyme successfully inside the right common carotid artery of rabbits[73].

#### *3.3.3.3. Mechanisms of action of papain*

Papain is a cysteine-proteinase type of endolytic enzyme extracted from the latex of green papaya (*Carica papaya*). It weighs 23,000 Da , and its molecules form a single peptide chain with 211 amino acid residues that fold into two distinct parts, which are divided by a cleft that represents its active site[74]. In addition to papain, the latex contains three additional enzymes (chymopapain, caricain, and glycil endopeptidase), which together with papain represent 80% of the enzymatic fraction, where papain corresponds to the smallest enzymatic fraction (5-8%). Although purification of papain is usually performed using precipitation techniques, it remains contaminated by other proteases[75]. With regard to its enzymatic activity, papain is activated by the addition of substances such as cyanide, reduced glutathione, and sulfate and is inactivated by oxidants. The maximal enzymatic activity occurs with a pH between 5 and 7.5. With regard to its specificity, in addition to hydrolyzing several substances, papain exhibits strong esterase activity, which makes its scope of action even wider to the point of acting on the very same substrates as pancreatic proteolytic enzymes with esterase activity[76].

Regarding its biological effects, papain exhibits remarkable elastolytic properties and has been successfully used in the production of experimental lung emphysema in animals[77, 78]. In addition to digesting elastic fibers, papain is also able to destroy collagen. Junqueira (1980) studied the ability of papain to destroy the collagen fibers of several tissues (cartilage, bone, skin, and blood vessel) from several animal species, such as *Gallus gallus* (chicken), *Canis familiaris* (dog), *Oryctolagus cuniculus* (rabbit) and *Sus scrofa* (pig), and observed that the degree of collagen destruction varies according to the type of tissue[79]. Ionescu (1977) used papain to de-antigenize a venous heterograft to subsequently graft it onto the common carotid artery of dogs and observed that papain caused an excessive weakening of the graft with a tendency to form venous aneurysms. To overcome this problem, the author subjected the grafts to a previous treatment with formol to maintain their rigidity and flexibility and not form aneurysms[80].

With regard to commercial presentation, papain is found as raw latex (~ 12 U/mg), lyophilized powder (10 U/mg) and aqueous suspension (16-40 U/mg)[81].

#### *3.3.3.4. Creation of papain-induced aneurysms*

The technique applied in the construction of papain-induced aneurysms uses the right common carotid artery of rabbits and is fully surgical, based on the study by Hoh et al. However, this technique does not use angiography during the puncture of the right common carotid artery and injection of the enzyme. This simplification proved to be safe and efficacious, and no animal exhibited complications due to the unduly passage of papain to an aberrant vascular branch that was accidentally present in the neck of the animals. Other innovations were the removal of the aortic arch and the supra-aortic trunks, direct measurement of the macroscopic dimensions of aneurysms and vessels using a caliper, and quantitative histological studies by means of histomorphometry in addition to a qualitative histological analysis[60,73].

#### *3.3.3.5. Morphologic and geometric features*

Papain-induced aneurysms exhibited a size similar to the elastase-induced aneurysms described in previous studies. Nevertheless, it is noteworthy to stress that the papaininduced aneurysms were measured directly on the right common carotid artery. This is an important point because most of the studies performed using elastase employed digital subtraction angiography to measure the aneurysms, which led to an overestimate of the aneurysm size. Thus, if papain-induced aneurysms were also measured by means of digital subtraction angiography, then their size would have most likely been overestimated. Independent from the method used, papain was efficacious in producing saccular aneurysms with an average diameter of 3.8 +/- 1.4 mm (2.5-7.0 mm), similar to those appearing in the ophthalmic segment of the human internal carotid artery.

#### *3.3.3.6. Histology*

58 Aneurysm

*3.3.3.3. Mechanisms of action of papain* 

proteolytic enzymes with esterase activity[76].

not form aneurysms[80].

*3.3.3.4. Creation of papain-induced aneurysms* 

of rabbits; however, that model exhibited a higher tendency to spontaneously rupture[72]. To produce saccular aneurysms larger than 5 mm, De Oliveira et al. (2011) infused the

Papain is a cysteine-proteinase type of endolytic enzyme extracted from the latex of green papaya (*Carica papaya*). It weighs 23,000 Da , and its molecules form a single peptide chain with 211 amino acid residues that fold into two distinct parts, which are divided by a cleft that represents its active site[74]. In addition to papain, the latex contains three additional enzymes (chymopapain, caricain, and glycil endopeptidase), which together with papain represent 80% of the enzymatic fraction, where papain corresponds to the smallest enzymatic fraction (5-8%). Although purification of papain is usually performed using precipitation techniques, it remains contaminated by other proteases[75]. With regard to its enzymatic activity, papain is activated by the addition of substances such as cyanide, reduced glutathione, and sulfate and is inactivated by oxidants. The maximal enzymatic activity occurs with a pH between 5 and 7.5. With regard to its specificity, in addition to hydrolyzing several substances, papain exhibits strong esterase activity, which makes its scope of action even wider to the point of acting on the very same substrates as pancreatic

Regarding its biological effects, papain exhibits remarkable elastolytic properties and has been successfully used in the production of experimental lung emphysema in animals[77, 78]. In addition to digesting elastic fibers, papain is also able to destroy collagen. Junqueira (1980) studied the ability of papain to destroy the collagen fibers of several tissues (cartilage, bone, skin, and blood vessel) from several animal species, such as *Gallus gallus* (chicken), *Canis familiaris* (dog), *Oryctolagus cuniculus* (rabbit) and *Sus scrofa* (pig), and observed that the degree of collagen destruction varies according to the type of tissue[79]. Ionescu (1977) used papain to de-antigenize a venous heterograft to subsequently graft it onto the common carotid artery of dogs and observed that papain caused an excessive weakening of the graft with a tendency to form venous aneurysms. To overcome this problem, the author subjected the grafts to a previous treatment with formol to maintain their rigidity and flexibility and

With regard to commercial presentation, papain is found as raw latex (~ 12 U/mg),

The technique applied in the construction of papain-induced aneurysms uses the right common carotid artery of rabbits and is fully surgical, based on the study by Hoh et al. However, this technique does not use angiography during the puncture of the right common carotid artery and injection of the enzyme. This simplification proved to be safe and efficacious, and no animal exhibited complications due to the unduly passage of papain to an aberrant vascular branch that was accidentally present in the neck of the animals. Other innovations were the removal of the aortic arch and the supra-aortic trunks, direct measurement of the macroscopic dimensions of aneurysms and vessels using a caliper, and

lyophilized powder (10 U/mg) and aqueous suspension (16-40 U/mg)[81].

papain enzyme successfully inside the right common carotid artery of rabbits[73].

From a histological perspective, papain caused the destruction of elastic fibers, endothelial damage, thrombosis, and intimal fibrosis. These alterations are similar to those found in elastase-induced aneurysms, in which the only difference is the degree of thrombosis, which was more remarkable in the papain-induced aneurysms[73].

#### *3.3.3.7. Future of the enzymatic model*

Currently, there are no ideal animal models of experimental saccular aneurysms available. From a practical perspective, it is impossible for one single model to reproduce the full histological, geometric, and hemodynamic characteristics of the wide variety of aneurysms and human-related conditions. Nevertheless, the enzymatic model has been increasingly used in the production of saccular aneurysms due to its simplicity, easy execution, and lower cost, resulting from the use of small animals such as rabbits, in addition to allowing the control of height, width, and size of the aneurysm neck. Furthermore, the enzymatic model can be improved, as a wide variety of enzymes have not yet been tested. Despite the advantages of the enzymatic model, the use of both elastase and papain exhibits some limitations, such as an intramural inflammatory response, endothelial damage, and thrombosis. Indeed, thrombosis is the most important effect because it hinders the interpretation of the results of the embolization materials tested. However, even when they are present, the intra-aneurysmal thrombi do not invalidate this experimental model because under actual clinical conditions, most human aneurysms have thrombi present. Therefore, although they are not ideal for preclinical tests of embolization materials, enzymatic models most closely mimic the actual clinical conditions and thus exhibit a high potential to contribute to the study of the physiopathology of human intracranial aneurysms and testing of embolization materials and endovascular devices.

#### **Author details**

Ivanilson Alves de Oliveira *Neuroradiology, Experimental Medicine Laboratory, Universidade Federal de Sergipe-UFS, Brazil* 

#### **4. References**


[18] Fagundes D.J., Taha M.O. Modelo animal de doença:critérios de escolha e espécies de animais de uso corrente. Acta Cirúrgica Brasileira, 2004. 19(1): p. 59-65.

60 Aneurysm

**4. References** 

187-91.

Neurol, 1975. 3(1): p. 7-14.

Neurosurg, 1991. 75(1): p. 8-14.

1992. 34(1): p. 30-2.

1980. 53(5): p. 633-41.

Neurosurg, 1978. 48(3): p. 369-82.

J Roentgenol, 1999. 173(1): p. 133-8.

lesions. J Neurosurg, 1990. 73(6): p. 859-63.

Neurosurg Psychiatry, 1993. 56(9): p. 947-59.

[1] Nyström H.S.M. Development of intracranial aneurysms as revealed by electron

[4] Barker C.S. Peripheral cerebral aneurysm associated with a glioma. Neuroradiology,

[5] Frazee J.G., Cahan L.D. and Winter J. Bacterial intracranial aneurysms. J Neurosurg,

[6] Lee K.S., Liu S.S., Spetzler R.F., Rekate H.L. Intracranial mycotic aneurysm in an infant:

[7] Bohmfalk G.L., Story J.L., Wissinger J.P., Brown, W.E. Bacterial intracranial aneurysm. J

[8] Brown R.D., Jr., Wiebers D.O. and Forbes G.S. Unruptured intracranial aneurysms and arteriovenous malformations: frequency of intracranial hemorrhage and relationship of

[9] Brown B.M. and Soldevilla F. MR angiography and surgery for unruptured familial intracranial aneurysms in persons with a family history of cerebral aneurysms. AJR Am

[10] Schievink W.I., Parisi J.E, Piepgras D.G., Michels V.V. Intracranial aneurysms in Marfan's syndrome: an autopsy study. Neurosurgery, 1997. 41(4): p. 866-70; discussion 871. [11] Heiserman J.E., Drayer B.P., Fram E.K., Keller P.J. MR angiography of cervical

[12] Kopitnik T.A. and Samson D.S. Management of subarachnoid haemorrhage. J Neurol

[13] Dix G.A., Gordon W., Kaufmann A.M., Sutherland I.A., Sutherland G.R. Ruptured and unruptured intracranial aneurysms--surgical outcome. Can J Neurol Sci, 1995. 22(3): p.

[14] Winn H.R., Almaani W.S., Berga S.L., Jane J.A., Richardson A.E. The long-term outcome in patients with multiple aneurysms. Incidence of late hemorrhage and implications for

[15] Yasargil M.G. and Fox J.L. The microsurgical approach to intracranial aneurysms. Surg

[16] Guglielmi G., Vinuela F., Dion J., Duckwiler G. Electrothrombosis of saccular aneurysms via endovascular approach. Part 2: Preliminary clinical experience. J

[17] Derdeyn C.P., Barr J.D., Berenstein A., Connors J.J., Dion J.E., Duckwiler G.R., Higashida R.T., Strother C.M., Tomsick T.A., Turski P. The International Subarachnoid Aneurysm Trial (ISAT): a position statement from the Executive Committee of the American Society of Interventional and Therapeutic Neuroradiology and the American

Society of Neuroradiology. AJNR Am J Neuroradiol, 2003. 24(7): p. 1404-8.

fibromuscular dysplasia. AJNR Am J Neuroradiol, 1992. 13(5): p. 1454-7.

treatment of incidental aneurysms. J Neurosurg, 1983. 59(4): p. 642-51.

[2] Rhoton A.L., Jr. Anatomy of saccular aneurysms. Surg Neurol, 1980. 14(1): p. 59-66. [3] Senegor M. Traumatic pericallosal aneurysm in a patient with no major trauma. Case

microscopy. J Neurosurg, 1963. 20: p. 329-337.

report of a case. Neurosurgery, 1990. 26(1): p. 129-33.

report. J Neurosurg, 1991. 75(3): p. 475-7.


[51] Stehbens W.E. Evaluation of aneurysm models, particularly of the aorta and cerebral arteries. Exp Mol Pathol, 1999. 67(1): p. 1-14.

62 Aneurysm

Neurosurg, 1996. 85(3): p. 488-95.

Radiology, 1999. 213(1): p. 217-22.

Neuroradiology, 2003. 45(7): p. 430-4.

Roentgenol, 2000. 174(2): p. 349-54.

Radiology, 1965. 84: p. 531-5.

68(2): p. 216-27.

Rec, 1964. 148: p. 611-4.

Louis: Mosby; 1974. p. 2763-86.

the rabbit. Acta Anat, 1974. 87: p. 283-300.

rabbits. Surg Neurol, 1998. 49(5): p. 491-4.

vivo study. Radiology, 1998. 207(2): p. 519-23.

devices. AJNR Am J Neuroradiol, 2004. 25(8): p. 1387-90.

[36] Spetzger U., Reul J., Weis J., Bertalanffy H., Thron A., Gilsbach J.M. Microsurgically produced bifurcation aneurysms in a rabbit model for endovascular coil embolization. J

[37] Spetzger U., Reul J., Weis J., Bertalanffy H., Gilsbach J.M. Endovascular coil embolization of microsurgically produced experimental bifurcation aneurysms in

[38] Raymond J., Salazkin I., Metcalfe A., Robledo O., Gevry G., Roy D., Weill A., Guilbert F. Lingual artery bifurcation aneurysms for training and evaluation of neurovascular

[39] Kallmes D.F., Helms G.A., Hudson S.B., Altes T.A., Do H.M., Mandell J.W., Cloft H.J. Histologic evaluation of platinum coil embolization in an aneurysm model in rabbits.

[40] Kallmes D.F., Williams A.D., Cloft H.J., Lopes M.B., Hankins G.R., Helm G.A. Platinum coil-mediated implantation of growth factor-secreting endovascular tissue grafts: an in

[41] Hans F.J., Krings T., Möller-Hartmann W., Thiex R., Pfeffer J., Cherer K., Brunn A., Dreeskamp H., Stein K.P., Meetz A., Gilsbach J.M., Thron A. Endovascular treatment of experimentally induced aneurysms in rabbits using stents: a feasibility study.

[42] Struffert T., Roth C., Romeike B., Grunwald I.O., Reith W. Onyx in an experimental aneurysm model: histological and angiographic results. J Neurosurg, 2008. 109(1): p. 77-82. [43] Altes T.A., Cloft H.J., Short J.G., DeGast A., Do H.M., Helm G.A., Kallmes D.F. 1999 ARRS Executive Council Award. Creation of saccular aneurysms in the rabbit: a model suitable for testing endovascular devices. American Roentgen Ray Society. AJR Am J

[44] Short J.G., Fujiwara N.H., Marx W.F., Helm G.A., Cloft H.J., Kallmes D.F. Elastaseinduced saccular aneurysms in rabbits: comparison of geometric features with those of

[45] Adams D.F., Olin T.B. and Redman H.C. Catheterization of Arteries in the Rabbit.

[46] Angel-James J. Variations in the vasculature of the aortic arch and its major branches in

[47] Bugge J. Arterial supply of the cervical viscera in the rabbit. Acta Anat (Basel), 1967.

[48] Blanding J.D., Jr., Ogilvie R.W., Hoffman C.L., Knisely W.H. The Gross Morphology of the Arterial Supply to the Trachea, Primary Bronchi, and Esophagus of the Rabbit. Anat

[49] Du Boulay G.H. Comparative neuroradiologic vascular anatomy of experimental animals. In: Newton D.G. and Potts T.H. (ed.) Radiology of the skull and brain. Saint

[50] Chungcharoen D., Daly M.B.; Neil E. The efect of occlusion upon the intrasinusal presure with special reference to vascular communications between the carotid and

vertebral circulations in the dog, cat, and rabbit. J Physiol, 1952. 117: p. 56-76.

human aneurysms. JNR Am J Neuroradiol, 2001. 22(10): p. 1833-7.

