Animal Models of Focal Ischemia

Approaches to Inducing Focal Ischemia

The most common subtype of human stroke is occlusion of the middle cerebral artery (MCA) or its distal branches and thus the most relevant and widely used animal models involve different means of MCA occlusion (MCAO). The most common and reproducible approaches to induce MCAO are either through a small craniotomy to expose distal branches of the MCA (distal MCAO) that are ligated, clipped, or cauterized, or via insertion of a filament intraarterially into the internal carotid artery that is advanced to the bifurcation of the MCA (filament MCAO). Although distal MCAO is often used as a permanent ischemia model, reperfusion can be induced by temporary ligatures or thrombolysis in an embolic model. The disadvantage of the distal MCAO model is the need for craniotomy that can damage structures including the eye, temporalis muscle, and zygomatic arch. The filament MCAO model avoids opening the skull, but is not without complications. Subarachnoid hemorrhage can be easily induced if filaments are too sharp or advanced too far. Filaments coated with silicone help to avoid this complication and also allow for different filament dimensions that are helpful to adjust to the size or strain of animal. The advantage of the filament model is that restoration of blood flow is readily achieved by removing the filament. This is particularly relevant to modeling endovascular therapy with stent retrievers, now a standard of care for certain patients with large-vessel occlusions.

Local injection of the vasoconstrictor endothelin has also been used to induce focal ischemia but the injection requires a craniotomy for placement of a small cannula. However, the cannula can be left in place and ischemia induced after anesthesia is withdrawn. The disadvantage of using endothelin is that the sensitivity to vasoconstriction varies depending on species, strain, sex, and comorbidity, as well as the presence or absence of anesthesia, making control of ischemia and timing of reperfusion difficult and variable.

Several embolic and thrombotic models have been developed to better mimic a common mechanism of occlusion and/or to study thrombolysis. Embolic models typically involve either injection of a suspension of small clots or careful intraarterial placement of a large clot into the MCA. These embolic models are more difficult to control ischemia and have high variability in infarct size. Thromboembolic models can induce direct thrombus formation and involve injection of thrombin or ferric chloride, or use a photothrombotic dye that is activated by certain wavelength light through a thinned skull or craniotomy into either proximal MCA or distal MCA branches. Although thromboembolic models better mimic the mechanism of human stroke, the timing of reperfusion can be variable and uncertain. Although thromboembolic models are useful for studying the effects of thrombolysis and tissue plasminogen activator (tPA) as a stroke treatment, it should be noted that different species have different fibrinolytic systems and thus may not be as relevant for human stroke .

In general, the approaches listed here can and have been used in different species, large and small, including mice, rats, cats, dogs, pigs, and nonhuman primates. The benefit of small rodents is that they are readily available and cost-effective, with more available reagents than larger species, and easier to use especially if a craniotomy is needed. The disadvantage of using rodents is that their lissencephalic brain is not as complex as the gyrencephalic brain of humans and primates and, therefore, the complexity of stroke pathology is different. Other differences in vascular anatomy and cellular responses likely confound translation from rodents to humans. For example, the anatomy of leptomeningeal collaterals has been shown to vary over ∼15 different mouse strains that impacts the size of infarction . However, the advantages of using rodents in most studies generally outweigh the disadvantages.

Induction of Middle Cerebral Artery Occlusion: Considerations for Anesthesia and Monitoring Physiological Variables

In general, a surgical approach is needed to induce MCAO. Thus experience with appropriate surgical techniques, including sterile conditions, use of anesthesia, and controlling and monitoring physiological variables, is critical to reproducibility and reliability of results.

Although a few models induce focal ischemia without anesthesia (e.g., endothelin occlusion, clot injection), it is generally impractical to perform these models on unanesthetized animals. Thus understanding the effects of anesthesia on factors that influence the ischemic cascade or induction of ischemia is important. Ketamine and barbiturates are intrinsically neuroprotective and are not recommended for use in stroke models. Chloral hydrate is an injectable anesthesia that does not appreciably affect blood flow or depress neuronal function , making it the preferred choice for studies involving measuring seizure or periinfarct depolarizations. The disadvantage to chloral hydrate is it can lower blood pressure, even in hypertensive animals, and requires an intravenous (IV) line. Inhalational anesthetics such as isoflurane are preferred in stroke models if not measuring brain excitability because the depth of anesthesia is easily controlled and animals recover quickly. However, isoflurane is a cerebral vasodilator that can make inducing ischemia difficult if the level is not kept at <2%. Inhalation anesthetics are also toxic and thus should be contained through ventilation or trapping. Laboratory personnel should be monitored for exposure. A combination of anesthesia can also be used that involves initial use of isoflurane for instrumentation (intubation, placement of flow probes, catheters, EEG leads, etc.) followed by tapering off isoflurane and onto an injectable anesthesia, such as chloral hydrate. Care should be taken to not overanesthetize during the change of anesthesia, but this approach is relatively simple since isoflurane is rapidly eliminated. Animals should be continuously monitored to maintain appropriate surgical anesthesia including periodic testing for lack of eye blink or toe pinch reflex, slow deep breathing, and increased blood pressure indicating pain.

Physiological parameters including blood gases (Pa o ₂ and Pa co ₂ ), pH, blood pressure, and temperature are important to monitor and potentially control during MCAO surgeries because they affect stroke outcome. CO ₂ levels affect cerebral blood flow (CBF) through vasodilation or vasoconstriction and thus maintaining Pa co ₂ within the physiological range of 35–45 mm Hg will eliminate this as a variable to outcome. Similarly, O ₂ levels can influence blood flow and the pathophysiology of stroke and thus should be ∼90–100 mm Hg. Blood gases can be monitored through periodic arterial blood sampling (suggest at least one before induction of MCAO and once during MCAO) in rats; however, mice are a challenge because of their low blood volume. Pulse oximetry and CO ₂ exhalation can also be used if blood samples are not feasible. Focal ischemia also causes significant fluctuations in ions, including Na ⁺ , K ⁺ , and H ⁺ due to ion channel inactivation during hypoxia/ischemia that can affect outcome. Thus controlling arterial pH to ∼7.4 can eliminate this as a variable. If surgeries are prolonged (>1 h), intubation and mechanical ventilation to control blood gases and pH are recommended. Arterial blood pressure is also known to influence stroke outcome. If blood pressure falls too low, stroke outcome is worsened. Because anesthesia can lower blood pressure, it should be continuously monitored through an arterial catheter and adjusted with phenylephrine or norepinephrine if needed. This is particularly relevant when using hypertensive animals that may be normotensive during MCAO surgery with the use of certain anesthesia. Temperature also affects stroke outcome. Brain hypothermia is highly neuroprotective and since anesthesia can decrease body temperature, it is important to monitor and maintain core body temperature with a heating pad or lamp. However, there are cases in which monitoring physiological parameter without adjusting them is appropriate. For example, some neuroprotective pharmacologic agents are effective at reducing stroke severity due to promoting hypothermia (e.g., NBQX (2,3-dihydroxy-6-nitro-7-sulfamoyl-benzo[f]quinoxaline-2,3-dione) and MgSO ₄ ) . In contrast, some agents may cause hyperthermia and would, therefore, be contraindicated for stroke therapy.