NEUROSURGICAL FOCUS Neurosurg Focus 47 (1):E20, 2019

In vivo cerebral models

John W. Thompson, PhD,1,3 Omar Elwardany, MD,1,3 David J. McCarthy, MS,1,3 Dallas L. Sheinberg, BS,1,3 Carlos M. Alvarez, MD,1,3 Ahmed Nada, MD,1,3 Brian M. Snelling, MD,1,3,4 Stephanie H. Chen, MD,1,3 Samir Sur, MD,1,3 and Robert M. Starke, MD1–3

Departments of 1Neurological Surgery and 2Radiology, University of Miami; 3The University of Miami Cerebrovascular Initiative, University of Miami; and 4Marcus Neuroscience Institute, Boca Raton Regional Hospital, Boca Raton, Florida

Cerebral aneurysm rupture is a devastating event resulting in and is associated with signifi- cant morbidity and death. Up to 50% of individuals do not survive aneurysm rupture, with the majority of survivors suf- fering some degree of neurological deficit. Therefore, prior to aneurysm rupture, a large number of diagnosed patients are treated either microsurgically via clipping or endovascularly to prevent aneurysm filling. With the advancement of endovascular surgical techniques and devices, endovascular treatment of cerebral is becoming the first-line therapy at many hospitals. Despite this fact, a large number of endovascularly treated patients will have aneurysm re- canalization and progression and will require retreatment. The lack of approved pharmacological interventions for cere- bral aneurysms and the need for retreatment have led to a growing interest in understanding the molecular, cellular, and physiological determinants of cerebral aneurysm pathogenesis, maturation, and rupture. To this end, the use of animal cerebral aneurysm models has contributed significantly to our current understanding of cerebral aneurysm biology and to the development of and training in endovascular devices. This review summarizes the small and large animal models of cerebral aneurysm that are being used to explore the pathophysiology of cerebral aneurysms, as well as the develop- ment of novel endovascular devices for aneurysm treatment. https://thejns.org/doi/abs/10.3171/2019.4.FOCUS19219 KEYWORDS aneurysm; animal; model; in vivo; mice; rabbit; porcine; canine

nruptured cerebral aneurysms (CAs) are common pathology of CAs. Equally vital to our understanding of in the general population, with an estimated prev- CA biology and treatment has been the use of CA animal alence ranging from 2% to 6%.68 If left untreated, models, which attempt to replicate the morphological, his- aneurysmsU can progress and spontaneously rupture, pro- tological, and hemodynamic features observed in human ducing a subarachnoid hemorrhage and resulting in sig- CAs. These animal models provide a method for inves- nificant morbidity and death. The pathophysiology of CA tigating aneurysm formation, growth, and rupture while formation and rupture is not fully defined, but risk factors also providing a means of testing new treatment modali- have been identified including increasing age, female sex, ties. CA models have been developed in numerous species , excessive alcohol intake, and .16,34,68 including mice, rats, rabbits, swine, sheep, canines, and Studies have suggested that hemodynamic stress is a criti- primates, with each model having advantages and limita- cal factor in CA pathogenesis17 leading to endothelial dys- tions such that the model selection depends on the purpose function, inflammatory cell infiltration, and arterial wall of the study. This review explores some of the more com- remodeling.6–7 Vascular smooth-muscle cells undergo a monly used models of CAs and compares the advantages phenotypic switch, which exacerbates by ex- and disadvantages of each system. pressing inflammatory and matrix remodeling proteins,50 ultimately culminating in histological changes character- ized by disruption of the internal elastic lamina, extracel- Small Animal CA Models lular matrix digestion, thinning of the media, cell loss, and The theory behind CA formation in rats and mice is aneurysm formation. that weakening of the cerebral blood vessels combined Molecular and histological analysis of human CA spec- with hemodynamic stress will induce CA formation. Nu- imens has revealed significant information regarding the merous rat and mouse models of CA formation exist and

ABBREVIATIONS BAPN = b-aminopropionitrile; CA = cerebral aneurysm; CCA = common carotid ; DOCA = deoxycorticosterone acetate. SUBMITTED March 1, 2019. ACCEPTED April 9, 2019. INCLUDE WHEN CITING DOI: 10.3171/2019.4.FOCUS19219.

©AANS 2019, except where prohibited by US copyright law Neurosurg Focus Volume 47 • July 2019 1

Unauthenticated | Downloaded 09/25/21 09:34 PM UTC Thompson et al.

FIG. 1. Cerebral aneurysm formation in rodents, hemodynamic stress, and vessel wall weakening. The procedures used for CA formations in rats and mice vary primarily in the method of inducing hypertension, increasing the flow rate, and weakening the ves- sel wall. Hypertension can be induced by a combination of a high salt diet, unilateral nephrectomy or bilateral ligation of the posterior branches of the renal (not shown), and subcutaneous placement of DOCA pellets or II–filled microosmotic pump (not shown). Increases in flow rate are accomplished by ligation of the left CCA, which causes a compensatory increase in flow rate in the contralateral . Vessel wall weakening is accomplished by feeding a diet containing 0.12% BAPN, a lysyl oxidase inhibitor, or by a single stereotactic injection of elastase. Copyright Robert Starke. Published with permission. primarily differ in the mechanisms of vessel wall weaken- at the bifurcation of the right anterior cerebral artery and ing and hemodynamic stress induction (Fig. 1). the olfactory artery. Histological analysis revealed frag- mented elastic laminin and media thinning suggestive of Hemodynamic Stress and Vessel Wall Weakening aneurysm formation in 78% of the treated mice. However, Hemodynamic stress in the cerebral vasculature can be the CAs formed by this method are small with a few mi- increased by hypertension and/or an increase in flow rate. croaneurysms observable by light microscopy, while other Using the combination of hypertension and flow rate to aneurysms require electron microscopy for visualization. induce hemodynamic stress, Hashimoto et al. created the This method of CA formation suffers from slow aneurysm first rodent CA model in rats.31 During a series of surger- formation. Other adaptations to this protocol include liga- ies, hemodynamic stress was increased by ligation of the tion of the left renal artery, unilateral nephrectomy, and bilateral ligation of the posterior branches of the renal ar- left common carotid artery (CCA) while hypertension was 3–6,8,10 induced by unilateral nephrectomy, followed by subcuta- teries during the same surgery. neous injections of deoxycorticosterone acetate (DOCA) and the addition of 1% sodium chloride to the drinking Elastase and Angiotensin II water. Vessel walls were weakened by feeding the rats Early stages of aneurysm formation are associated with chow containing 0.12% b-aminopropionitrile (BAPN), a elastic lamina degeneration, which may contribute to an- lysyl oxidase inhibitor, which prevents and elas- eurysm progression and rupture. Given this histological tin cross-linking, leading to increased vessel fragility and finding, Nuki et al.54 stereotactically injected elastase into a greater likelihood of aneurysm formation. Morimoto et the of the right basal cistern. To induce al. later adapted this method for CA formation in mice and hypertension, angiotensin II was continuously adminis- included bilateral ligation of the posterior branches of the tered via a subcutaneously placed microosmotic pump. renal arteries.51 Four months following surgery, CAs were CA formation was achieved in 77% of the mice within 2 observed at various stages of formation, located primarily weeks of treatment. Histologically, the aneurysms demon-

2 Neurosurg Focus Volume 47 • July 2019

Unauthenticated | Downloaded 09/25/21 09:34 PM UTC Thompson et al. strated degeneration of the media layer and elastic lamina tigation of aneurysm biology at a molecular, cellular, and and infiltration of inflammatory cells. physiological level. Excluding surgically created saccular aneurysms, rodent CA models do not require direct vessel Rupture Model manipulation and have an intracranial location. This leads The spontaneous aneurysm rupture model was in- to the question of what constitutes an aneurysm. In the troduced by Makino et al.,41 who used a combination of early studies, aneurysm formation produced small micro- elastase treatment to weaken cerebral blood vessels and aneurysms that were rarely visible and only detectable by hypertension. With this method, in a series of surger- light or electron microscopy or histological alterations of ies, hypertension is induced by unilateral nephrectomy, the vessel wall. Some scientists do not believe these “mi- implantation of a DOCA-salt pellet, and the addition of croaneurysms” recapitulate human CA disease. In con- 1% NaCl to the drinking water. During the same surgery trast, elastase treatment results in clear and defined out- ward bulging of the vessel walls of the circle of Willis and as the DOCA-salt pellet implantation, the mice receive 61 a single injection of elastase into the right basal cistern. its major branches. Starke et al. defined an aneurysm as a bulge in the vessel wall whose diameter is > 150% of the With this method, CA formation occurs in > 60% of the 54 mice within 28 days of the aneurysm induction surgery. diameter of the parent artery. Similarly, Nuki et al. de- Additionally, spontaneous aneurysm rupture occurs in fined an aneurysm as a bulging of the vessel wall > 150% 50%–60% of mice within 7–11 days following surgery. of the diameter of the basilar artery. Hosaka et al. later modified this model with the addition Although rodent CA models replicate many of the his- of an increased vessel flow rate and fragility induced by tological and molecular changes found in human CAs, ligating the left CCA and the right renal artery, followed 1 there are certain pathologies observed in human CAs but week later by the injection of elastase into the right basal not in the rodent CA models. For example, in human sac- cistern.33 Hypertension and vessel fragility were further cular CAs, lipids and oxidized lipids accumulate in the aneurysm wall and are associated with cell death, vessel enhanced by angiotensin II and by chow containing 8% 22,55 NaCl and 0.12% BAPN. Using this method with elastase wall weakening, and aneurysm rupture. Similarly, the concentrations greater than 50 mU, 100% of the mice de- complement inflammatory system is activated in human saccular CAs and is involved in aneurysm wall degrada- velop CAs. 65 With elastase, CAs are made and also rupture at pre- tion and rupture. dictable time points. With 25–30 mU of elastase, the ma- The commercial availability of genetically modified jority of mice form aneurysms at 1 week without signs of mice has made rodent CA models a vital tool in investi- rupture.61–64 However, approximately 80% of animals will gating the molecular underpinnings of CA formation, pro- have subarachnoid hemorrhage by 4 weeks. Similar to the gression, and rupture. Transgenic mice allow for the in- histological changes observed in human CAs, the aneu- vestigation of particular proteins that are altered in human rysms formed by elastase display disruption of the elastic CA disease. For example, tumor factor (TNF)– a,61 chemoattractant protein (MCP)–1,4 and lamina, infiltration, loss or reduction of the k 7 , and smooth-muscle cell hyperplasia. This nuclear factor (NF)– B p50 subunit knockout reduces CA formation, whereas endothelial model is utilized extensively in the literature and has been 9 40 used to test pharmacological inhibitors that decrease the (eNOS) and SOX17 knockout predisposes mice to CA of aneurysm progression and rupture.54,62 formation. The cited studies are just a few of the many using transgenic mice to better understand CA biology in Surgically Created Saccular Aneurysms preclinical studies. The small animal size and the intracranial aneurysms formed preclude the use of rodent CA models for endo- Large Animal CA Models vascular device testing. To circumvent this issue, Frösen Large animal CA models have been made in numer- et al.21 and Marbacher et al.43 surgically created saccular ous species but are primarily formed in rabbits, dogs, and aneurysms using a donor thoracic aorta, which was surgi- swine.15 Aneurysm formation in large animals requires di- cally ligated end-to-side to the abdominal aorta in both rect vessel manipulation through either microsurgical or mice and rats. These saccular aneurysms display inflam- endovascular intervention, and these aneurysms are typi- matory cell infiltration, endothelial denudation, cally formed using the CCA. Therefore, these models are formation, and intimal hyperplasia. Marbacher et al.44 extracranial in location and suffer from the effects of sur- expanded on this model with sodium dodecyl sulfate–in- gical creation at the aneurysm neck and dome.15 Despite duced decellularization of the donor thoracic aorta prior these weaknesses, each model has particular characteris- to aneurysm creation. The loss of mural cells led to an un- tics that are either advantageous for or detrimental to the organized luminal thrombus, increased inflammation, and purposes of a particular study. wall damage resulting in aneurysm growth and rupture.44 Although mice are too small, studies have been success- Rabbit Aneurysm Models fully conducted using the rat saccular aneurysm model for Venous Graft Aneurysm 11,27 45 testing stents and coils. To simulate arterial bifurcation aneurysms, a technique was developed to create venous pouch aneurysms using Perspectives and Limitations a jugular venous graft at a surgically induced bifurca- Rodent CA models offer a powerful tool for the inves- tion at the end-to-side anastomosis of the left CCA to the

Neurosurg Focus Volume 47 • July 2019 3

Unauthenticated | Downloaded 09/25/21 09:34 PM UTC Thompson et al.

to the CCA ligation to prevent retrograde elastase flow.38 Finally, to dissolve the collagen in the , colla- genase was added to the elastase in endovascular models, which resulted in aneurysms that are nearly histologically identical to human CAs, showing fragmentation and dimi- nution of the internal elastic lamina and increased levels of smooth-muscle cells.36,37,44,45 Although aneurysms cre- ated with this method show long-term patency more than 24 months, these aneurysm fail to grow and rupture and demonstrate a homogeneity in vessel wall makeup and thickness. Therefore, it fails to replicate the more com- plex, heterogenic aneurysm vessel environment of athero- sclerosis and wall thinning, as well as the inflammatory cell infiltration and de-endothelialization associated with aneurysm rupture. Hemodynamic Stress–Induced Aneurysms of the Posterior Circulation Studies by Hassler in 196332 and later by Gao et al.23 demonstrated that aneurysms can be formed in the pos- FIG. 2. Rabbit elastase aneurysm model. Aneurysm formation in rabbits consists of exposing the right CCA. After gaining arterial access, a bal- terior circulation of rabbits using hemodynamic stress loon is advanced to the origin of the CCA (A). As the balloon is inflated, alone without hypertension or vessel wall weakening. In elastase is simultaneously injected, filling the artery (B). The artery is this model, hemodynamic stress is increased in the basilar incubated with elastase for 20 minutes. The elastase and balloon are artery by unilateral or bilateral ligation of the carotid ar- then removed, and the distal portion of the CCA is ligated, forming the teries. Using this technique, Hassler and Gao found histo- aneurysm. Residual elastase and hemodynamic forces will cause the logical changes in the arterial wall of the basilar terminus aneurysm to maturate over a period of several weeks following surgery. resembling nascent aneurysm formation characterized by A digital subtraction study shows a newly formed aneurysm (C). Copyright Robert Starke. Published with permission. a loss of internal elastic lamina, media thinning, and an outward bulge of the vessel . This model has been expanded with the addition of the aneurysm risk factors of hypertension and estrogen deficiency.66 Hypertension right CCA.20 Marbacher et al.42 modified this technique is induced by unilateral nephrectomy combined with a to create a more complicated aneurysm by grafting large, high salt diet, and estrogen deficiency is induced by bi- wide-necked, bilobar, and bisaccular venous pouches to lateral oophorectomy. The combining of hemodynamic this surgically created bifurcation. Most recently, a very stress with hypertension and estrogen deficiency induced broad-necked aneurysm model was developed by longitu- changes in the circle of Willis, such as vessel length and dinally opening a segment of the jugular and graft- 59 tortuosity, as well as aneurysm lesion formation and vas- ing this patch to the CCA. A limitation of these venous cular damage. pouch models is their inability to replicate the histological changes observed in human aneurysms. A beneficial qual- Fusiform Aneurysm ity of this model is the ability to tailor aneurysms in terms Recently, Avery et al.12 developed a carotid artery fu- of both dome and neck size. siform aneurysm in rabbits. In this model, the right CCA Arterial Aneurysms is exposed and wrapped in gauze and isolated from sur- rounding tissue by placing the CCA-wrapped section into The most common model of aneurysm creation in rab- a cradle. The gauze is then soaked in elastase and CaCl2 bits is the saccular aneurysm created by enzymatic weak- for 20 minutes. With this method, fusiform aneurysms, ening of the arterial wall. A reliable method was devel- which were defined as vessel dilations greater than 50% oped whereby aneurysms were produced by occluding of the proximal artery diameter, were formed in 100% of the CCA with an endovascular balloon and incubating the the animals at 6 weeks after aneurysm creation surgery. aortic arch–brachiocephalic trunk bifurcation with elas- 18 Histologically, these aneurysms demonstrate an almost tase. Subsequently, the same group refined their tech- complete loss of the internal elastic lamina, a reduction nique whereby the CCA was surgically exposed, cannulat- in the tunica media, and a thickening of the . ed, and occluded with an endovascular balloon, and then The long-term patency of this model was not investigated the CCA was incubated with elastase above the balloon past 6 weeks. prior to its distal ligation (Fig. 2).1 While early techniques for endovascular incubation with elastase were effective Canine Aneurysm Models at producing aneurysms with adequate patency duration, the retrograde flow of elastase would often cause damage Venous Pouch Model to the trachea and other organs. To curtail this complica- The first reliable aneurysm model was developed in tion, the technique was modified further so that a micro- 1954 by German and Black, who used a venous pouch was introduced distal to the balloon and proximal graft to create saccular aneurysms in dogs.24 This tech-

4 Neurosurg Focus Volume 47 • July 2019

Unauthenticated | Downloaded 09/25/21 09:34 PM UTC Thompson et al.

FIG. 3. Venous pouch aneurysms. Illustration depicting the surgical creation of sidewall, bifurcation, and terminal aneurysms using the left and right CCA and a segment of the external jugular vein (EJV). Copyright Robert Starke. Published with permission.

nique has remained in use to date.62,70 The technique modified by using a longer venous segment and a side- involves exposing the external jugular vein and section- to-side anastomosis to construct giant aneurysms, which ing a suitable length; one side of the venous segment is were more prone to rupture if left untreated than smaller closed with a suture to create a venous pouch, which is sized aneurysms.56 then sutured to an arteriotomy created at any location of Elastase was introduced by Goericke et al.25 to create the investigator’s choice. Several modifications of this saccular aneurysms. As in rabbit models, the CCA is ex- technique have been described including those for gi- posed and occluded at the origin, and elastase is injected ant, wide-necked, and fusiform aneurysm creation (Fig. into the CCA and incubated for a period of time. As in 3).29,35,​57,​71,72 both the rodent and rabbit CA models, elastase weakens Generally, the CCA or cervical internal carotid artery the arterial wall, triggering an initial inflammatory re- is selected for the creation of aneurysms because of their sponse as well as activation of endogenous proteinases similarity in caliber and blood flow to human cerebral to break down and collagen, resulting in vascular vessels and the ability of the animal to tolerate the surgi- dilation.30 cal procedure.26,47,60,69 The canine’s CCA is approximately 4 mm in diameter, similar to the human internal carotid Perspectives and Limitations artery, and the relatively long CCA in dogs (10–12 cm) Large animal CA models offer broad utility for inves- affords easy surgical access. tigating endovascular therapeutic interventions, healing, Hemodynamic Stress and Arterial Wall Injury and endovascular training. Among the large animal mod- els, the venous pouch aneurysm model allows for the se- More recently, Wang et al. described a novel method of lection of aneurysm size, morphology, and location, and CA formation by inducing hemodynamic stress in com- 67 the aneurysm can be created in vessels with a caliber and bination with arterial wall weakening. In this model, a blood flow similar to those of human cerebral vessels. new branch in the CCA is surgically constructed by at- However, the venous pouch model suffers in terms of the taching the proximal segment of one CCA to the proximal 49,61,67 surgical trauma and suture material involved in aneurysm sidewall of the contralateral CCA. Hypertension is formation, nonarterial aneurysm composition, and an arti- induced, and elastase is delivered externally to the apex of ficial neck. Despite these drawbacks, large animal models the newly created bifurcation. allow for testing of endovascular devices as well as endo- vascular training. However, the preferred CA model for Swine Aneurysm Models endovascular training is still up for debate. The procedure for aneurysm production in swine is A major disadvantage of the rabbit elastase aneurysm similar to that described by German and Black for an- model is that it lacks an inflammatory response and does eurysm formation in canines.24 This method was slightly not spontaneously rupture, but it does have coagulation and

Neurosurg Focus Volume 47 • July 2019 5

Unauthenticated | Downloaded 09/25/21 09:34 PM UTC Thompson et al. thrombolysis profiles similar to those of humans, which is Acknowledgments critical for testing new materials for use in endovascular 13 This work was supported by a National Research and Educa- devices for aneurysm occlusion. A major disadvantage of tion Foundation (NREF) Young Clinician Investigator Award, Joe the swine CA venous graft model is a tendency for spon- Niekro Research Grant, Bee Foundation Award, Aneurysm taneous and healing with or without emboli- Foundation Award, and Miami Clinical and Translational Science zation.28,39,​52,58 Another disadvantage of the large animal Institute Award (R.M.S.). The project described was supported models is the presence of viable mural vascular smooth- by grant no. UL1TR002736, Miami Clinical and Translational muscle cells, which are significantly reduced or absent in Science Institute, from the National Center for Advancing Trans- 44 lational Sciences and the National Institute on Minority Health human CA tissue. Marbacher et al. demonstrated that and Health Disparities. Its contents are solely the responsibility of decellularized aneurysm grafts formed an unorganized the authors and do not necessarily represent the official views of luminal thrombus and had increased inflammation and the NIH. wall damage resulting in aneurysm growth and rupture. Therefore, the healing response in CA models with normal References cellularization of the aneurysm wall would be enhanced and thereby potentially enhance the healing response in 1. Altes TA, Cloft HJ, Short JG, DeGast A, Do HM, Helm GA, et al: Creation of saccular aneurysms in the rabbit: a model device studies. suitable for testing endovascular devices. AJR Am J Roent- genol 174:349–354, 2000 Silicone Aneurysm Models 2. Amili O, Schiavazzi D, Moen S, Jagadeesan B, Van de Moor- tele PF, Coletti F: Hemodynamics in a giant intracranial The recent advancement in and accessibility to 3D aneurysm characterized by in vitro 4D flow MRI. PLoS One printed technologies has allowed for the fabrication of pa- 13:e0188323, 2018 tient-specific true-to-scale arterial replicas.19,37,46 These 3D 3. Aoki T, Fukuda M, Nishimura M, Nozaki K, Narumiya S: printed models serve as education tools for presurgical as- Critical role of TNF-alpha-TNFR1 signaling in intracranial sessment or can be further processed using silicone-cast- aneurysm formation. Acta Neuropathol Commun 2:34, ing technology to form a hollow, silicone-walled artificial 2014 4. Aoki T, Kataoka H, Ishibashi R, Nozaki K, Egashira K, vasculature. There is an increasing volume of literature in Hashimoto N: Impact of monocyte chemoattractant protein-1 which artificial CA models have been used for endovascu- deficiency on cerebral aneurysm formation. 40:942– lar device testing and training, surgical clip ligation train- 951, 2009 ing, presurgical assessment, and fluid dynamics studies,2 5. Aoki T, Kataoka H, Ishibashi R, Nozaki K, Hashimoto N: and they have also been surgically implanted into swine Cathepsin B, K, and S are expressed in cerebral aneurysms and cadaveric human heads for neurosurgical training.14,53 and promote the progression of cerebral aneurysms. Stroke Although this model offers an excellent alternative to ani- 39:2603–2610, 2008 mals in endovascular training, it does not fully replicate 6. Aoki T, Kataoka H, Moriwaki T, Nozaki K, Hashimoto N: Role of TIMP-1 and TIMP-2 in the progression of cerebral the natural arterial biology, which may greatly affect test- aneurysms. Stroke 38:2337–2345, 2007 ing results. 7. Aoki T, Kataoka H, Shimamura M, Nakagami H, Wakayama K, Moriwaki T, et al: NF-kB is a key mediator of cerebral Clinical Translation and Conclusions aneurysm formation. Circulation 116:2830–2840, 2007 8. Aoki T, Nishimura M, Kataoka H, Ishibashi R, Nozaki K, Animal models of CA have been and continue to be an Hashimoto N: modulate growth of invaluable tool for investigating the molecular, cellular, and cerebral aneurysms: a study using the free scavenger physiological aspects of CA pathophysiology as well as for edaravone and p47phox-/- mice. Lab Invest 89:730–741, 2009 testing novel endovascular devices. Ideally, the CA model 9. Aoki T, Nishimura M, Kataoka H, Ishibashi R, Nozaki K, will replicate the hemodynamic forces, wall sheer stresses, Miyamoto S: Complementary inhibition of cerebral aneu- rysm formation by eNOS and nNOS. Lab Invest 91:619– and cellular and tissue responses observed in human CAs. 626, 2011 However, no animal model perfectly replicates the human 10. Aoki T, Nishimura M, Matsuoka T, Yamamoto K, Furuyas- disease being investigated. Therefore, each investigator hiki T, Kataoka H, et al: PGE2-EP2 signalling in endothelium must consider the strengths and weaknesses of each model is activated by haemodynamic stress and induces cerebral in order to best replicate the aspect of CAs that is being aneurysm through an amplifying loop via NF-kB. Br J investigated. In general, rodent CA models are useful for Pharmacol 163:1237–1249, 2011 investigating the molecular and cellular mechanisms of 11. Aquarius R, Smits D, Gounis MJ, Leenders WPJ, de Vries J: implantation in a rat model of sidewall an- aneurysm formation, growth, and rupture with the goal eurysm: a feasibility study. J Neurointerv Surg 10:88–92, of finding druggable targets for therapeutic intervention 2018 and translational potential. In contrast, large CA animal 12. Avery MB, Alaqeel A, Bromley AB, Chen YX, Wong JH, models are primarily used in the development and refine- Eesa M, et al: A refined experimental model of fusiform an- ment of new endovascular therapies and in the assess- eurysms in a rabbit carotid artery. J Neurosurg 131:88–95, ment of novel therapeutic interventions, as was done with 2019 Gamma Knife radiosurgery.48 Large animal models also 13. Bavinzski G, al-Schameri A, Killer M, Schwendenwein I, allow for the investigation of aneurysm healing following Gruber A, Saringer W, et al: Experimental bifurcation aneu- rysm: a model for in vivo evaluation of endovascular tech- therapy. No current CA model perfectly replicates human niques. Minim Invasive Neurosurg 41:129–132, 1998 CA disease. Therefore, further work is needed to create a 14. Benet A, Plata-Bello J, Abla AA, Acevedo-Bolton G, Saloner CA model that more closely replicates the histological and D, Lawton MT: Implantation of 3D-printed patient-specific pathophysiological features of human CA disease. aneurysm models into cadaveric specimens: a new training

6 Neurosurg Focus Volume 47 • July 2019

Unauthenticated | Downloaded 09/25/21 09:34 PM UTC Thompson et al.

paradigm to allow for improvements in cerebrovascular sur- Investigators: Unruptured intracranial aneurysms—risk of gery and research. BioMed Res Int 2015:939387, 2015 rupture and risks of surgical intervention. N Engl J Med 15. Bouzeghrane F, Naggara O, Kallmes DF, Berenstein A, Ray- 339:1725–1733, 1998 mond J: In vivo experimental intracranial aneurysm models: 35. Jiang YZ, Lan Q, Wang QH, Wang SZ, Lu H, Wu WJ: Cre- a systematic review. AJNR Am J Neuroradiol 31:418–423, ation of experimental aneurysms at a surgically created 2010 arterial confluence. Eur Rev Med Pharmacol Sci 19:4241– 16. Brinjikji W, Zhu YQ, Lanzino G, Cloft HJ, Murad MH, Wang 4248, 2015 Z, et al: Risk factors for growth of intracranial aneurysms: a 36. Kang W, Connor J, Yan X, Neely B, Carney E, Ellwanger J, systematic review and meta-analysis. AJNR Am J Neurora- et al: A modified technique improved histology similarity diol 37:615–620, 2016 to human intracranial aneurysm in rabbit aneurysm model. 17. Chalouhi N, Hoh BL, Hasan D: Review of cerebral aneurysm Neuroradiol J 23:616–621, 2010 formation, growth, and rupture. Stroke 44:3613–3622, 2013 37. Kimura T, Morita A, Nishimura K, Aiyama H, Itoh H, Fu- 18. Cloft HJ, Altes TA, Marx WF, Raible RJ, Hudson SB, Helm kaya S, et al: Simulation of and training for cerebral aneu- GA, et al: Endovascular creation of an in vivo bifurcation rysm clipping with 3-dimensional models. aneurysm model in rabbits. Radiology 213:223–228, 1999 65:719–726, 2009 19. D’Urso PS, Thompson RG, Atkinson RL, Weidmann MJ, 38. Krings T, Möller-Hartmann W, Hans FJ, Thiex R, Brunn A, Redmond MJ, Hall BI, et al: Cerebrovascular biomodelling: a Scherer K, et al: A refined method for creating saccular aneu- technical note. Surg Neurol 52:490–500, 1999 rysms in the rabbit. Neuroradiology 45:423–429, 2003 20. Forrest MD, O’Reilly GV: Production of experimental aneu- 39. Lee D, Yuki I, Murayama Y, Chiang A, Nishimura I, Vinters rysms at a surgically created arterial bifurcation. AJNR Am HV, et al: Thrombus organization and healing in the swine J Neuroradiol 10:400–402, 1989 experimental aneurysm model. Part I. A histological and mo- 21. Frösen J, Marjamaa J, Myllärniemi M, Abo-Ramadan U, lecular analysis. J Neurosurg 107:94–108, 2007 Tulamo R, Niemelä M, et al: Contribution of mural and bone 40. Lee S, Kim IK, Ahn JS, Woo DC, Kim ST, Song S, et al: marrow-derived neointimal cells to thrombus organization Deficiency of endothelium-specific transcription factor Sox17 and wall remodeling in a microsurgical murine saccular an- induces intracranial aneurysm. Circulation 131:995–1005, eurysm model. Neurosurgery 58:936–944, 2006 2015 22. Frösen J, Tulamo R, Heikura T, Sammalkorpi S, Niemelä M, 41. Makino H, Tada Y, Wada K, Liang EI, Chang M, Mobashery Hernesniemi J, et al: Lipid accumulation, lipid oxidation, and S, et al: Pharmacological stabilization of intracranial aneu- low plasma levels of acquired antibodies against oxidized rysms in mice: a feasibility study. Stroke 43:2450–2456, lipids associate with degeneration and rupture of the intracra- 2012 nial aneurysm wall. Acta Neuropathol Commun 1:71, 2013 42. Marbacher S, Erhardt S, Schläppi JA, Coluccia D, Remonda 23. Gao L, Hoi Y, Swartz DD, Kolega J, Siddiqui A, Meng H: L, Fandino J, et al: Complex bilobular, bisaccular, and Nascent aneurysm formation at the basilar terminus induced broad-neck microsurgical aneurysm formation in the rabbit by hemodynamics. Stroke 39:2085–2090, 2008 bifurcation model for the study of upcoming endovascular 24. German WJ, Black SP: Experimental production of carotid techniques. AJNR Am J Neuroradiol 32:772–777, 2011 aneurysms. N Engl J Med 250:104–106, 1954 43. Marbacher S, Frösén J, Marjamaa J, Anisimov A, Honkanen 25. Goericke SL, Parohl N, Albert J, Dudda M, Forsting M: P, von Gunten M, et al: Intraluminal cell transplantation Elastase-induced aneurysm in swine: proof of feasibility in a prevents growth and rupture in a model of rupture-prone sac- first case. A technical note. Interv Neuroradiol 15:413–416, cular aneurysms. Stroke 45:3684–3690, 2014 2009 44. Marbacher S, Marjamaa J, Bradacova K, von Gunten M, 26. Graves VB, Ahuja A, Strother CM, Rappe AH: Canine model Honkanen P, Abo-Ramadan U, et al: Loss of mural cells of terminal arterial aneurysm. AJNR Am J Neuroradiol leads to wall degeneration, aneurysm growth, and eventual 14:801–803, 1993 rupture in a rat aneurysm model. Stroke 45:248–254, 2014 27. Grüter BE, Täschler D, Strange F, Rey J, von Gunten M, 45. Marjamaa J, Tulamo R, Frösen J, Abo-Ramadan U, Hernesni- Grandgirard D, et al: Testing bioresorbable stent feasibility emi JA, Niemelä MR, et al: Occlusion of neck remnant in in a rat aneurysm model. J Neurointerv Surg [epub ahead of experimental rat aneurysms after treatment with platinum- print], 2019 or polyglycolic-polylactic acid-coated coils. Surg Neurol 28. Guglielmi G, Ji C, Massoud TF, Kurata A, Lownie SP, Vi- 71:458–465, 2009 ñuela F, et al: Experimental saccular aneurysms. II. A new 46. Mashiko T, Otani K, Kawano R, Konno T, Kaneko N, Ito Y, model in swine. Neuroradiology 36:547–550, 1994 et al: Development of three-dimensional hollow elastic model 29. Guo S, Jiang P, Liu J, Yang X, Jiang C, Li Y, et al: A com- for cerebral aneurysm clipping simulation enabling rapid and parative CFD analysis of common carotid fusiform aneurysm low cost prototyping. World Neurosurg 83:351–361, 2015 in canine models and vertebrobasilar fusiform aneurysm in 47. Massoud TF, Guglielmi G, Ji C, Viñuela F, Duckwiler GR: human patients. Int Angiol 37:32–40, 2018 Experimental saccular aneurysms. I. Review of surgically- 30. Halpern VJ, Nackman GB, Gandhi RH, Irizarry E, Scholes constructed models and their laboratory applications. Neuro- JV, Ramey WG, et al: The elastase infusion model of experi- radiology 36:537–546, 1994 mental aortic aneurysms: synchrony of induction of endog- 48. Meadowcroft MD, Cooper TK, Rupprecht S, Wright TC, enous proteinases with matrix destruction and inflammatory Neely EE, Ferenci M, et al: Gamma Knife radiosurgery cell response. J Vasc Surg 20:51–60, 1994 of saccular aneurysms in a rabbit model. J Neurosurg 31. Hashimoto N, Handa H, Hazama F: Experimentally induced 129:1530–1540, 2018 cerebral aneurysms in rats. Surg Neurol 10:3–8, 1978 49. Meng H, Wang Z, Hoi Y, Gao L, Metaxa E, Swartz DD, et al: 32. Hassler O: experimental carotid ligation followed by aneurys- Complex hemodynamics at the apex of an arterial bifurcation mal formation and other morphological changes in the circle induces vascular remodeling resembling cerebral aneurysm of Willis. J Neurosurg 20:1–7, 1963 initiation. Stroke 38:1924–1931, 2007 33. Hosaka K, Downes DP, Nowicki KW, Hoh BL: Modified mu- 50. Mérei FT, Gallyas F: Role of the structural elements of the rine intracranial aneurysm model: aneurysm formation and arterial wall in the formation and growth of intracranial sac- rupture by elastase and hypertension. J Neurointerv Surg cular aneurysms. Neurol Res 2:283–303, 1980 6:474–479, 2014 51. Morimoto M, Miyamoto S, Mizoguchi A, Kume N, Kita T, 34. International Study of Unruptured Intracranial Aneurysms Hashimoto N: Mouse model of cerebral aneurysm: experi-

Neurosurg Focus Volume 47 • July 2019 7

Unauthenticated | Downloaded 09/25/21 09:34 PM UTC Thompson et al.

mental induction by renal hypertension and local hemody- 64. Tada Y, Wada K, Shimada K, Makino H, Liang EI, Muraka- namic changes. Stroke 33:1911–1915, 2002 mi S, et al: Estrogen protects against intracranial aneurysm 52. Murayama Y, Viñuela F, Suzuki Y, Akiba Y, Ulihoa A, rupture in ovariectomized mice. Hypertension 63:1339– Duckwiler GR, et al: Development of the biologically active 1344, 2014 Guglielmi detachable coil for the treatment of cerebral an- 65. Tulamo R, Frösen J, Junnikkala S, Paetau A, Pitkäniemi J, eurysms. Part II: an experimental study in a swine aneurysm Kangasniemi M, et al: Complement activation associates with model. AJNR Am J Neuroradiol 20:1992–1999, 1999 saccular cerebral artery aneurysm wall degeneration and rup- 53. Namba K, Mashio K, Kawamura Y, Higaki A, Nemoto S: ture. Neurosurgery 59:1069–1077, 2006 Swine hybrid aneurysm model for endovascular surgery 66. Tutino VM, Mandelbaum M, Takahashi A, Pope LC, Sid- training. Interv Neuroradiol 19:153–158, 2013 diqui A, Kolega J, et al: Hypertension and estrogen deficiency 54. Nuki Y, Tsou TL, Kurihara C, Kanematsu M, Kanematsu Y, augment aneurysmal remodeling in the rabbit circle of Wil- Hashimoto T: Elastase-induced intracranial aneurysms in lis in response to carotid ligation. Anat Rec (Hoboken) hypertensive mice. Hypertension 54:1337–1344, 2009 298:1903–1910, 2015 55. Ollikainen E, Tulamo R, Lehti S, Lee-Rueckert M, Hernesni- 67. Wang J, Tan HQ, Zhu YQ, Li MH, Li ZZ, Yan L, et al: Com- emi J, Niemelä M, et al: cell forma- plex hemodynamic insult in combination with wall degen- tion, apolipoproteins, and ABCA1 in intracranial aneurysms: eration at the apex of an arterial bifurcation contributes to implications for lipid accumulation as a promoter of aneu- generation of nascent aneurysms in a canine model. AJNR rysm wall rupture. J Neuropathol Exp Neurol 75:689–699, Am J Neuroradiol 35:1805–1812, 2014 2016 68. Weir B: Unruptured intracranial aneurysms: a review. J Neu- 56. Raymond J, Darsaut TE, Kotowski M, Makoyeva A, Gevry rosurg 96:3–42, 2002 G, Berthelet F, et al: Thrombosis heralding aneurysmal rup- 69. Yan L, Zhu YQ, Li MH, Tan HQ, Cheng YS: Geometric, he- ture: an exploration of potential mechanisms in a novel giant modynamic, and pathological study of a distal internal ca- swine aneurysm model. AJNR Am J Neuroradiol 34:346– rotid artery aneurysm model in dogs. Stroke 44:2926–2929, 353, 2013 2013 57. Raymond J, Salazkin I, Georganos S, Guilbert F, Desfaits 70. Yang XJ, Li L, Wu ZX: A novel arterial pouch model of AC, Gevry G, et al: Endovascular treatment of experimental saccular aneurysm by concomitant elastase and collagenase wide neck aneurysms: comparison of results using coils or digestion. J Zhejiang Univ Sci B 8:697–703, 2007 cyanoacrylate with the assistance of an aneurysm neck bridge 71. Yapor W, Jafar J, Crowell RM: One-stage construction of device. AJNR Am J Neuroradiol 23:1710–1716, 2002 giant experimental aneurysms in dogs. Surg Neurol 36:426– 58. Raymond J, Venne D, Allas S, Roy D, Oliva VL, Denbow 430, 1991 N, et al: Healing mechanisms in experimental aneurysms. 72. Ysuda R, Strother CM, Aagaard-Kienitz B, Pulfer K, Con- I. cells and neointima formation. J signy D: A large and giant bifurcation aneurysm model in Neuroradiol 26:7–20, 1999 canines: proof of feasibility. AJNR Am J Neuroradiol 59. Sherif C, Herbich E, Plasenzotti R, Bergmeister H, Wind- 33:507–512, 2012 berger U, Mach G, et al: Very large and giant microsurgical bifurcation aneurysms in rabbits: Proof of feasibility and comparability using computational fluid dynamics and bio- Disclosures mechanical testing. J Neurosci Methods 268:7–13, 2016 60. Sorteberg A, Sorteberg W, Rappe A, Strother CM: Effect of Dr. Snelling has direct stock ownership in RIST Neurovascular. Guglielmi detachable coils on intraaneurysmal flow: experi- mental study in canines. AJNR Am J Neuroradiol 23:288– Author Contributions 294, 2002 Conception and design: Starke, Thompson. Drafting the article: 61. Starke RM, Chalouhi N, Jabbour PM, Tjoumakaris SI, Gon- Thompson, Elwardany, McCarthy, Sheinberg, Alvarez, Nada, zalez LF, Rosenwasser RH, et al: Critical role of TNF-a in Snelling, Chen, Sur. cerebral aneurysm formation and progression to rupture. J Neuroinflammation 11:77, 2014 Correspondence 62. Starke RM, Thompson JW, Ali MS, Pascale CL, Martinez Robert M. Starke: Miami University, Lois Pope LIFE Center, Lege A, Ding D, et al: Cigarette smoke initiates oxidative Miami, FL. [email protected]. stress-induced cellular phenotypic modulation leading to cerebral aneurysm pathogenesis. Arterioscler Thromb Vasc Biol 38:610–621, 2018 63. Tada Y, Wada K, Shimada K, Makino H, Liang EI, Muraka- mi S, et al: Roles of hypertension in the rupture of intracra- nial aneurysms. Stroke 45:579–586, 2014

8 Neurosurg Focus Volume 47 • July 2019

Unauthenticated | Downloaded 09/25/21 09:34 PM UTC