Intraluminal Suture (Monofilament) Middle CerebralArteryOcclusion Model
The intraluminal suture (monofilament) MCAO model is the most commonly used in focal brain ischemia studies, probably because it is relatively easy to perform and less invasive than other models. Because a craniectomy is not required, damage from brain retraction, vessel manipulation, temperature loss, and desiccation of exposed brain are all avoided. In this model, the MCA is occluded by inserting a monofilament suture into the internal carotid artery (ICA) to block blood flow to the MCA (Fig. 1). Either permanent or transient MCAO can be simply achieved by maintaining or withdrawing the monofilament suture. This suture MCAO model was originally described in rats in 1986 and later modified (26). Usually, a 3-0 or 4-0 monofilament suture is used as the occluder. The monofilament can be coated with silicone (26,27) or poly-L-lysine (28), or it can be used without coating (29,30). The suture occluder can be inserted through the CCA, the ICA, or the external carotid artery. The length of suture inserted from the bifurcation of the CCA is approximately 17 to 22 mm (31-33), depending on body weight, size of the suture tip, rat strain, and location of the bifurcation.
The typical infarct areas induced by this suture model include both the lateral caudoputamen and frontoparietal cortex (Fig. 2). It is established that a substantial ischemic penumbra exists in this model early after ischemia (34) and that the infarct size induced by prolonged ischemia (>90 min) is relatively reproducible, thus making this model appropriate for testing neuroprotective agents. However, it must be remembered that many factors can affect infarct size. First, one study showed that a slight difference in aspects of the monofilament (i.e., diameter, tensile strength, and extensibility) can cause a significant difference in infarct volume (35 ) . Second, the ischemia induced by a silicone-coated suture is likely more substantial and less variable (28,29); consequently, the infarct volume is more reproducible and larger than that induced by an uncoated suture ( 29,30 ). Third, a longer insertion distance of the monofilament suture might give rise to larger infarction because the deeper insertion of the suture can also obstruct blood flow in some branches of the anterior cerebral artery (32). Lastly, inadvertent premature reperfusion might also cause variability of infarct volume (36). It is therefore important that consistent and standardized surgical procedures and techniques be used to generate reproducible lesions.
The suture MCAO method was also successfully applied to mice because transgenic or knockout mice provide unique avenues for basic research into the molecular mechanisms that
contribute to ischemic cell damage and development of novel therapeutic interventions (37-39). Usually, a 5-0 monofilament nylon suture or 8-0 suture coated with silicone is used to occlude blood flow to the MCA territory. The insertion depth from the bifurcation of the CCA is approximately 9 to 11 mm (38,39). The lesion is reproducible and its distribution is similar to that of the rat. Recently, the suture occlusion model could successfully be applied to neonatal rats that weigh 14 to 19 g (40) and to adult rabbits (41).
Another important role for this MCAO method is that it can be modified to induce ischemia in an MRI unit by remotely advancing the suture occluder, the so-called "in-bore suture MCAO model" (33,42-44). Combined with new MRI techniques, this in-bore occlusion method enables researchers to monitor in vivo ischemic changes at a very early time point after the onset of ischemia and to acquire both pre- and postischemic data for later pixel-by-pixel comparison. The in-bore MCAO method was improved recently and achieved a high success rate (33 ).
Similar to the intraluminal suture occlusion model, some investigators used a silicone plastic catheter as a plug to occlude the MCA in dogs via the left ICA (45). In this study, MCAO and reperfusion upon withdrawal of the catheter could be demonstrated by angiography in all 19 mongrel dogs. Unfortunately, infarct sizes and locations varied largely (45). Various catheter models with inflating balloon tips have been used to occlude the MCA in several species.
Some disadvantages and complications are associated with the intraluminal suture MCAO model. Subarachnoid hemorrhage might occur because of inadvertent arterial rupture caused by the suture, a complication that is likely to have higher incidence with uncoated sutures than with silicone-coated sutures (28,36). In addition, spontaneous hyperthermia occurs when the ischemic duration lasts >2 hr, most likely due to ischemic damage of the hypothalamus, which is caused by blockage of small branches to the hypothalamus by suture insertion (46,47). Finally, the inner surface of vessels might be mechanically injured by the suture, which can complicate reperfusion.
Application of laser-Doppler flowmetry or electroencephalography during and after induction of ischemia in this model substantially reduced the incidence of complications, such as subarachnoid hemorrhage, incomplete MCAO, or premature reperfusion (36). Ipsilateral cortical laser-Doppler flowmetry signal immediately and abruptly decreases to approximately 20% of baseline following MCAO and substantially recovers following premature or induced reperfusion. Similarly, MCAO causes severe suppression of the ipsilateral hemispheric electroencephalographic signal, with a delay of 5 to 10 sec and recovers after reperfusion. Both methods could substantially increase successful MCAO rates and reduce the complications sometimes associated with this model (36 ).
Focal ischemia induced by thromboemboli is of great interest because of its resemblance to human ischemic stroke (approximately 80% of which are caused by thromboembolism) (48 ) and because of its role in evaluating thrombolytic therapy. Thrombolytic therapy, with recombinant tissue plasminogen activator (rtPA) administered intravenously within 3 hr after onset of ischemic stroke, improves neurologic outcome in humans (49) and is a standard therapy in carefully selected
Figure 2 Demonstration of infarction on T2-weighted MRI and with TTC staining from a rat subjected to the suture-occlusion model. Abbreviation: TTC, 2,3,5-triphenyltetrazolium chloride.
patients. Thromboembolic animal models are therefore playing an increasingly important role. Single-clot embolization models were developed for several species, including rabbit (50), dog (51,52), rat (53,54), mouse (55,56), guinea pig (57), and monkey (58,59).
Thromboembolic ischemia can be produced by a photochemical approach or by injection of autologous or heterologous thrombi. The photochemical method produces an arterial lesion in the CCA that results in formation of a platelet-rich thrombus, which can then dislodge and embolize to distal vessels (60,61). The photochemically induced thromboemboli are platelet-rich and, therefore, might not be amenable to thrombolytic therapy with rtPA.
The most commonly used thromboembolic model is blood-clot injection (Fig. 3) in the rat. In the early versions of this model, a suspension of microembolic clot was injected, causing diffuse and inhomogeneous infarction in the MCA territory because of peripheral branch microemboli-zation. Scattered, multifocal lesions were also observed in the territories of the anterior cerebral artery and posterior cerebral artery and even in the contralateral hemisphere (53,54). In addition, early spontaneous recanalization frequently occurred, which made evaluation of thrombolytic therapy difficult. The early autolysis of blood clots might be due to a more fragile red thrombus formed in vitro by whole blood. To overcome these problems, a more resistant white thrombus was produced by using a mobile, high-pressure, closed compartment system (PE-10 polyethylene tube, 0.28 mm in inner diameter) (62 ) . Using white thrombi, a substantial reduction of cerebral blood flow (CBF) was demonstrated in the affected region, with no spontaneous recan-alization at 2 hr after embolization, a condition necessary for studying thrombolytic treatment (62). However, infarct size was variable and ischemia caused by multiple small clots did not mimic typical clinical ischemic stroke. An ideal thromboembolic model should employ a blood clot that appropriately lodges in the proximal segment of the MCA, with the distal branches remaining open. Smaller clots embolize distally into the cerebral vessels, whereas larger clots lodge in vessels too proximal from the origin of MCA. Therefore, the size (length and diameter) and characteristics (i.e., more rigid fibrin-rich clot) of blood clots are crucial in this model. Recently (63), a rat clot model was developed in which 12 medium-sized (1.5 x 0.35 mm), fibrin-rich autologous clots formed in a PE-50 catheter (0.58 mm in inner diameter) were injected to produce reliable occlusion of the proximal MCA. Consistent reduction of CBF and histologic damage in the MCA territory were observed. Visual inspection demonstrated no early spontaneous clot lysis in the ipsilateral vessels and no clots in the contralateral vessels at 3 hr
after injection. It was demonstrated that thrombolytic therapy with rtPA (63) or prourokinase (64) can recanalize the occluded MCA. By inserting a modified tube into the ICA 2 to 3 mm proximal to the MCA origin, a single, fibrin-rich clot (25 x 0.1 mm) could also be selectively introduced into the proximal part of the MCA (65), or a thrombosis could be induced at the origin of the MCA (66), thus causing typical MCAO. Using this single-clot model, a significant reduction of CBF in the MCA territory was demonstrated, and the blood clot in the MCA trunk was found at 24 hr after embolization (66). This single clot was also applied to occlude the proximal MCA in mouse (55), but one disadvantage is the relatively high incidence of subarachnoid and intra-parenchymal hemorrhage. The very same model was successfully transferred to monkeys and led to the development of reproducible infarctions without hemorrhagic complications (59).
In conclusion, these single-clot or medium-sized, multiple-clot models induce predictable and reproducible infarcts in both extent and size in the MCA territory, similar to those caused by the intraluminal suture model. Embolization of the proximal MCA trunk by a single, fibrin-rich clot is similar to human embolic ischemia, because the majority of human ischemic strokes are caused by a single embolus in the MCA territory (1). Therefore, the single-clot model is promising for studying the pathogenesis of ischemic stroke and thrombolytic therapy.
Direct Surgical Middle Cerebral Artery Occlusion Model
Direct surgical MCAO is invasive, as it requires a craniectomy. In this model, ischemia can be induced by directly ligating, coagulating, clipping, or snaring the MCA trunk or its branches. The MCAO in the rat can be permanent or transient (67-69), can be performed under anesthesia or in awake animals and has been performed in nonhuman primates (70,71), dogs (25,72), rabbits (73,74), cats (75,76), rats (77), and mice (78). The rat is the most common species on which surgical MCAO has been performed.
Many models of focal brain ischemia have utilized temporary or permanent occlusion of the MCA in various animal species, with various routes described, including the subfrontal route (8 ) , the retro-orbital route (79 ), and the transorbital route (71,80 ) . With all of these large-animal models, the fundamental problem was the extreme variability of infarct size among animals. Among the first studies with reproducible infarct sizes was the description of a cat model with left MCAO via a transorbital approach, combined with bilateral ligation of internal carotid arteries (81). The transorbital method requires removal of the eyeball but is otherwise considered to be atraumatic, because it is not necessary to retract the brain for exposure of the vessel. The postorbital approach to the MCA in cats was developed later and revealed reproducible infarct sizes (82). The advantage of this model is that it does not necessitate enucleation of the eye, but it is technically much more complicated.
Direct ligation of the distal MCA at the rhinal fissure in the rat (77) was first described in 1975. The typical ischemic injury induced by such a method is in the frontoparietal cortex, but the extent of lesion is quite variable, ranging from 1 to 5 mm in diameter. Surgical occlusions more proximally in the MCA trunk were performed through a subtemporal craniectomy (83,84). Occluding the main trunk of the MCA proximal to the lenticulostriatal branches that supply the lateral caudoputa-men results in an infarction always involving both the frontal cortex and the lateral part of the caudoputamen. A more distal occlusion of the MCA (sparing the lenticulostriate branches) restricts the infarction to the cortex (68). Tandem occlusion of the MCA and the ipsilateral CCA improves the reproducibility of infarction (85 ). Similarly, simultaneous occlusions of the MCA and the ipsilateral CCA, combined with transient occlusion of the contralateral CCA, leads to good reproducibility of infarcts (86). Use of microclips, hooks, or ligature snares instead of electrocoagu-lation allows for later reperfusion (67,87,88). The infarction areas induced by a proximal MCAO appear to be larger and less variable, compared with those induced by distal MCAO. As focal occlusion of the MCA might not always produce infarction, even when the occlusion is performed at the proximal MCA trunk, researchers have refined this model and demonstrated that both the site and extent of MCAO affect the neuropathologic outcome and are critical factors in producing reproducible infarction (89). Occlusion of a very short segment (1-2 mm) of the MCA resulted in greater variability of infarction, and the rate of infarction was low when focal MCAO was performed at the origin of the MCA. This might relate to the persistence of abundant collateral circulation. Furthermore, because MCA branching is variable (90), such a focal occlusion at one point might not really involve both the lenticulostriatal and cortical branches. Interestingly, in the young (36 days) rat, MCAO beyond the point of origin of the lenticulostriate branches did not cause neuronal injury, probably because of better collateral supply (5 ).
A long, segmental occlusion of the MCA (3-6 mm) beginning proximal to the olfactory tract, however, induces uniform infarction in both extent and location in all rats. This extensive occlusion involves the lenticulostriate and small cortical branches and, therefore, produces reproducible infarction. However, the ischemic penumbra in this model might be small (91) and less amenable to testing neuroprotective compounds. This model is appropriate for investigating cerebrovascular function after focal ischemia (92). Because of higher long-term survival of rats due to the relatively small lesion, this model is a good option for screening stroke restorative drugs (7). Interestingly, in one study, MCAO alone led to infarction in only 50% of mongrel dogs, whereas 75% developed infarcts upon tandem occlusion of the MCA and the ipsilateral carotid artery (93 ).
Some disadvantages are associated with direct surgical MCAO models. First, performing this model is more difficult and requires more experience and technical skill due to varying MCA anatomic patterns (84). Second, direct exposure of brain to air after craniectomy might change intracranial pressure and blood-brain barrier permeability (5). Third, a small amount of subarachnoid hemorrhage around the MCA trunk can occur (84). Although these models had been extensively used, they have now been largely replaced by the intraluminal suture MCAO model.
Many materials, such as plastic microspheres, silicone cylinders, viscous silicone, collagen, and air, can be injected to induce ischemic injury as emboli via the CCA or the ICA (94-96). The magnitude and severity of ischemic damage induced by these embolic models depends on the number and size of the embolic particles injected (97). The injection of sodium arachidonate into the ICA of rats induced widespread intravascular cerebral thrombosis (98). Previously, diffuse distribution and inhomogeneous infarction were the neuropathologic hallmarks of these models, making histologic evaluation difficult. However, these models have recently been refined, and injection of 6 ceramic spheres, 0.3 to 0.4 mm in diameter, into the CCA via the external carotid artery leads to reproducible infarcts (99). The advantage of this model is the lack of potential hyperthermia, but the disadvantage is that it does not allow reperfusion. Single silver or golden ball injections into the ICA, which leads to MCAO, have already been used for decades in larger animals (100 ).
Various amounts of microfibrillar collagen, which was frequently used in neurosurgery as a topical hemostatic agent, were injected directly into the left CCA and resulted in cerebral infarcts of various sizes, shapes, locations (within the MCA territory), and number in 9 out of 10 mongrel dogs (101). A catheter with an inflatable balloon tip that was inserted into the MCA of baboons and was kept in place for 3 hr led to formation of an in situ thrombosis (102). Similarly, artificial embolic materials have successfully been used for inducing MCA-territory ischemia in several species, including monkeys (103 ).
In this model, the ischemic injury is induced by vascular injection of a photoactive dye, such as rose bengal (104) or photofrin (105), in combination with irradiation with a light beam at a specific wavelength. It has been shown that a reaction between the circulating dye and the light engenders free radicals, leading to platelet aggregation and thrombosis (104). The location and extent of photochemically induced lesions can be well controlled by selectively illuminating the brain tissue and by using different intensities of light and different doses of dye. A typical lesion in this model is a sharply circumscribed infarct that involves only the cortex. However, this model results in a small penumbral region. In addition, breakdown of the blood-brain barrier and vasogenic edema occur very early in this model (106,107). It is debatable whether the lesion induced by this model is secondary to an ischemic event. Although this model was used to test neuroprotective agents (108), its usefulness is limited, because it is believed that the pathophysiologic processes induced in this model are likely to be less relevant to those in human ischemic stroke.
Endothelin-Induced Middle Cerebral Artery Occlusion Model
Endothelin-1 is a 21-amino acid peptide with a potent vasoconstrictor effect (109). Application of endothelin-1 to the exposed MCA induces a significant decrease in CBF. Furthermore, microinjection of endothelin-1 into areas near the MCA through a cannula also decreases CBF in the
MCA territory. The distribution of ischemic infarct induced by this method is similar to that following permanent surgical ligation of the MCA (110). Interestingly, CBF around the isch-emic core significantly increases in this model. The advantage of this model is that ischemia can be induced in conscious rats, which excludes the confounding effects of general anesthesia. However, the ischemic damage is variable due to different responses of vessels to endothelin-1
Was this article helpful?