The Stroke Preclinical Assessment Network Multi-Laboratory Model of Thromboembolic Stroke with Thrombolysis: TE-MCAo

Summary

To simulate thromboembolic stroke, emboli prepared from heterologous rat blood were injected into the middle cerebral artery, followed by administration of systemic thrombolysis for recanalization. This version of the model is optimized for use in a multi-laboratory network and supports testing multiple candidate therapeutics.

Abstract

Current translational rodent stroke models induce middle cerebral artery occlusion (MCAo) using nylon filaments, injected emboli, intraluminal thrombin, or perivascular endothelin-1 to simulate human stroke. Among these methods, thromboemboli injection followed by thrombolysis best mimics the neuroinflammatory events seen in human patients and may be preferable to the inert nylon filament method most widely used. The standard thromboembolic models, used in leading single laboratories, however, can be time-consuming, produce variable results, and require considerable skill to master. To address these limitations, we developed a thromboembolic MCAo model that targets vessel occlusion and uses intravenous thrombolysis to achieve recanalization, paralleling the systemic thrombolysis administered in clinical scenarios. To reduce the number of animals, we developed a method to store blood from donor animals for later emboli preparation for multiple subjects. Our utilization of prefabricated Doccol microcatheters simplifies thrombus preparation and injection by pre-loading thromboemboli into the microcatheters that are then inserted into the internal carotid artery. To achieve recanalization, we infused intravenous Tenecteplase at a dose of 1.5 mg/kg. To promote reproducibility, we prepared and field-tested standard operating procedures, training videos, and hands-on surgical training workshops. The model uses a standard surgical approach that should be familiar to all investigators who use the widely accepted nylon filament model. Passage of the microcatheter into the distal internal carotid artery, avoiding the pterygopalatine artery, is accomplished in a manner similar to the nylon filament insertion. Recovery and post-stroke assessments may be done with typical behavioral, radiographic, and histologic protocols. This model offers a practical, reproducible, and accessible approach for investigators seeking a thromboembolic MCAo model with controlled recanalization.

Introduction

Stroke is a leading cause of death and disability worldwide¹. Many patients benefit from intravenous thrombolysis, and selected patients benefit from mechanical thrombectomy. Since not all patients arrive at a hospital in time for thrombolysis or thrombectomy, and since not all patients benefit from complete recovery, there remains a considerable need for further development of additional or adjunctive cerebroprotective therapeutics. The most widely used rodent ischemic stroke model is the monofilament occlusion of the middle cerebral artery (MCA) followed by complete reperfusion. This model simulates large vessel occlusion (LVO), modeling patients with large strokes undergoing successful complete mechanical thrombectomy. However, this model captures a small percentage of clinical strokes. In contrast, more patients suffer thromboembolic strokes secondary to atherosclerosis (in situ atherosclerotic plaque rupture or athero-emboli from systemic disease), and are treated with systemic thrombolysis alone, which the nylon model fails to replicate^2,3. In addition, the nylon filament MCAo model lacks the inflammatory cascade associated with thrombotic occlusion^4,5. Animal models employing thromboemboli followed by systemic thrombolytic administration would represent the pathophysiology and basic biology occurring in many stroke patients^4,6. So far, however, rodent thromboembolic stroke models are known to be complex and difficult to reproduce. In isolated laboratories with dedicated experts, a thromboembolic model can be achieved with reasonable reproducibility. Such a model does not readily lend itself to widespread use, as needed on multi-laboratory projects.

Herein, we provide detailed protocols for an updated and standardized thromboembolic middle cerebral artery occlusion (TE-MCAo) followed by systemic thrombolysis model. Specifically, we simplified the thrombus preparation, standardized thrombus-loading into catheters, and validated the scalability of this model for multi-laboratory trials. We developed this model for use in the Stroke Preclinical Assessment Network (SPAN). We intended, however, that the model be useful to any network seeking to implement high-volume pre-clinical network trials of new, candidate treatments. Using our standardized model, we were able to induce TE-MCAo into 6 rats per day per study site during the study period.

SPAN and similar networks have grown in response to repeated and ongoing failures to translate promising cerebroprotectants from the pre-clinical setting to success in clinical practice⁷. In 2019, the National Institutes of Neurological Disorders and Stroke (NINDS) initiated SPAN⁸. SPAN aims to more effectively screen potential therapeutics by minimizing bias through centralized masking, randomization, and automated centralized outcome analysis. SPAN 1 employed the nylon filament model in its first two iterations because all participating laboratories were already using it, and it was easy to deploy⁹. To more faithfully model many aspects of thromboembolism and thrombolysis, SPAN sought to create the TE-MCAo model described here. This model would be helpful for the preclinical surgeons who are transitioning from nylon filament model to TE-MCAo model. We recommend that preclinical labs should confirm the presence of stroke 24/48 h after MCAo (or TE-MCAo) using an imaging technique, e.g., magnetic resonance imaging (MRI).

The SPAN Network has been fully described in previous studies⁹, and all SPAN Standard Operating Procedures (SOPs) are available on the SPAN website. In brief, the network is managed by a coordinating center (CC) that handles drug distribution, data quality control, verification of protocol adherence, and network communication. All SOPs are drafted by the CC and approved by a Steering Committee after review. In SPAN, all rodents have an MRI compatible bar-coded ear tag placed upon arrival at the research laboratories, and all subjects are registered into the SPAN Research Electronic Data Capture (REDCap) Database. Animals are randomized by the CC using randomization tables stratified for site and sex. In SPAN 1 and SPAN 2, putative therapeutics were investigated for efficacy at six participating research laboratories. All test drugs were packaged in identically appearing vials, labeled at the coordinating center, and shipped to research laboratories at the start of the project.

Protocol

All participating SPAN laboratories, including the coordinating center at USC, followed NIH and AAALAC guidelines for the humane and ethical treatment of animals. Approval of the local IACUC was obtained at all sites.

1.Thrombus preparation

NOTE: All blood and thrombus should be handled according to all local institutional guidelines for biohazards. All materials that come into contact with blood should be either disposed of as biohazardous waste or, in the case of surgical instruments, thoroughly cleaned and disinfected by autoclave. To reduce the use of blood donor animals in this trial, we developed a method for storing donor blood and then using blood to make thromboemboli as needed.

1. Draw fresh donor blood from an artery into a 1.5 mL microcentrifuge tube or pediatric ethylenediamine tetra-acetic acid (EDTA) microtubes, filling only to the 500 µL level.
   NOTE: We prefer femoral artery cannulation for our blood draws, since the femoral artery has a large luminal diameter permitting catheterization, it is bilateral allowing for 2 separate blood draws, and is a survivable surgery for the rodents. Blood can be stored at 4 °C for 4 weeks, in our experience.
2. Add 1 mL of CaCl₂ solution (5mg/mL) to 500 µL of the stored, anticoagulated blood.
3. Immediately following anticoagulation reversal with CaCl₂, aspirate blood from the microcentrifuge tube into a 100 cm of PE-50 tubing using gentle suction.
   NOTE: Delay will result in blood clotting before it can be drawn into the PE-50 tubing.
4. Incubate the coiled tubing in pre-warmed phosphate-buffered saline (PBS) at 37 °C for 2 h in a table-top oven.
   NOTE: Placing the petri dish on an open hot plate results in an inconsistent result due to inconsistent heat distribution.
5. At the end of the 2 h incubation, immediately transfer the blood-filled PE-50 tubing to 4 °C for storage. It will remain usable for the next steps between 20-72 h.

2. Catheter loading

1. Connect the coiled PE-50 tubing (internal diameter: 0.58 mm) to a 3 or 5 cc syringe filled with saline and gently expel the thrombus into a petri dish containing PBS.
   NOTE: This may require considerable force. Hold back pressure on the connection site to ensure the catheter and syringe do not become disconnected upon extrusion. The PE-50 catheter can be cut into smaller segments (10-50 cm) if necessary to facilitate extrusion.
2. Cut sections of thrombus approximately 6 cm in length with a razor blade.
3. Wash each thrombus segment by drawing it gently into a PE-50 catheter and fully expelling it 5 times.
4. Next, wash the thrombus by drawing it gently into a PE-10 catheter (internal diameter: 0.279 mm) and fully expelling it 15 times.
   ​NOTE: If frequent breakage is experienced during this step, consider utilizing magnification with an operating microscope or a surgical loupe to ensure that the thrombus is being aspirated in line with the catheter opening.
5. OPTIONAL: Using Evan's Blue to visualize thrombi
   1. Prepare 4% Evan's Blue in 1x PBS solution.
   2. Add 200 µL of the 4% Evan's blue into a petri dish filled with 10 mL of 1x PBS. This will be the more concentrated solution = Evan's Blue Dilution # 1.
   3. Take 400 µL of Evan's Blue Dilution # 1 and mix into a separate petri dish filled with 20 mL of 1x PBS. This is a diluted solution of Evan's Blue = Evan's Blue Dilution # 2.
   4. Place the clot in the more concentrated Evan's Blue petri dish solution (dilution #1) for 1 s and then transfer to the diluted Evan's Blue petri dish (dilution #2).
6. Load one washed thrombus into prefabricated microcatheters (inner diameter: 0.2 mm, outer diameter: 0.360 mm, length: 300 mm, connected to female Luer lock) without air emboli using the wet-to-wet method.
   NOTE: We recommend using magnification (e.g., 2.5x loupes) for this step.
7. Once the thrombus is fully loaded, extrude enough thrombus so that exactly 5.0 cm remains in the microcatheter. Trim the excess with a razor blade.
   NOTE: The length of the loaded thrombus can be tailored to different applications depending on the size of the stroke desired.
8. Mark the microcatheter 16 mm from the tip, to demarcate the extent of the catheter that should be advanced intraluminally past the bifurcation of the common carotid artery.

3. Animal surgery and anesthesia

1. Induce anesthesia with 4% isoflurane in 30:70 oxygen:nitrous oxide mixture and maintain anesthesia with 1-2% isoflurane in the same gas mix.
2. Give analgesics carprofen 5 mg/kg subcutaneously and atropine 0.05 mg/kg intraperitoneally to maintain animal comfort intraoperatively.
   NOTE: Many surgeons infuse the incision site with 1.0 mL bupivacaine (2.5 mg/mL).
3. Maintain body temperature at 37 ± 0.5 °C throughout the surgery using a heating pad controlled with a rectal temperature probe.
4. Provide carprofen 5 mg/kg or meloxicam 1 mg/kg subcutaneously once or twice daily for at least 3 days as post-operative analgesia.

4. Collection of donor blood

1. Collect donor blood via femoral arterial catheterization. Match the biological sex of the donor and the recipients.
2. If blood is to be used on the same day:
   1. Prepare three 1.5 mL centrifuge tubes. Mark each tube at the 500 µL mark with an indelible laboratory pen.
   2. Prepare three sets of PE-50 tubing, each 100 cm long, attached to a 1.0 mL syringe via a 23-gauge blunt needle.
3. If blood is to be stored for later use, prepare six pediatric EDTA microtubes. Loosen the caps for easier use later.
4. Prepare a 10.0 cm length of PE-50 tubing, clamped at one end with a rubber-shod clamp and blunt-beveled at the other end.
   1. To make a rubber-shod clamp, insert one tine of a standard hemostatic clamp into PE-100 tubing and trim to fit. Repeat on the other side.
   2. To render the PE-50 blunt-beveled, first cut the end at a beveled angle. Then trim off the very tip of the bevel point using micro scissors.
      ​NOTE: It is important that the point be removed to avoid puncturing the back wall of the artery after insertion, but if too much of the tip is removed, the loss of the beveling effect will make it very difficult to insert the catheter intraluminally.
5. Perform a standard cut down onto the femoral artery.
6. Prepare the animal (shave inner thigh area), position in dorsal recumbency on surgery table. Sterilize the skin with 3 washes of 70% ethanol followed by 1 wash with betadine or other antiseptic solution. Disinfect the skin with 3 alternating rounds of betadine or other antiseptic scrub and 70% ethanol.
7. Secure hind legs in a splayed fashion using soft restraints.
8. Make a vertical incision on the inner thigh approximately 1.5-2 cm long along the natural fold of the inguinal intersection with the leg. This incision will be orthogonal to the expected anatomic course of the femoral artery.
9. Using blunt dissection, carefully dissect down to the femoral vessels. Carefully separate the femoral artery from the vein and nerve that run in the same bundle.
10. Distally occlude the femoral artery with a 5-0 silk ligature and use the tails of the suture to provide slight tension distally with hemostats.
11. Skeletonize the femoral artery as far proximal as permitted, tracing the artery to its retroperitoneal origin. Loosely place a loop of 5-0 silk around the femoral artery and use the tails again to provide gentle tension on the proximal end of the artery. Do not tighten the silk knot at this point. This tension will provide hemostasis after the arteriotomy is made.
    ​NOTE: The loop can be moved further proximally with an aneurysm clamp.
12. About 2/3 of the distance from the proximal suture loop, place a piece of silk suture under the artery and make a small arteriotomy distal to the loose suture using micro scissors. The loose suture will be tied down to secure the intraluminal catheter following cannulation.
13. Using sharp forceps, lift the top opening of the arteriotomy and introduce the blunt beveled catheter into the opening. Advance the catheter as far proximal as possible.
    ​NOTE: There may be friction between the catheter and the inside vessel wall. Two things can help with this: 1) coating the outside of the end of the catheter with serous fluid from the inside of the leg right before introducing it to the vessel, and 2) using blunt, grooved forceps, hold the outside of the vessel while advancing the catheter at the same time. This prevents the vessel from just advancing along with the catheter.
14. Once the tip of the catheter has been advanced almost to the point of the proximal loop, release the tension from the tails of that suture. Bleeding should be prevented by the tight fit of the catheter inside the vessel.
15. Advance the catheter enough so that the loose loop, when tightened, will encompass the catheter. Tighten the suture gently, ensuring that the suture is not tightened to the point of occluding the catheter. Also, ensure that the suture is not on the bevel of the catheter, given the tapered nature of this catheter segment.
16. Release the rubber-shod clamp. The catheter should immediately fill with blood.
17. Fill the open centrifuge tubes or the pediatric EDTA microtubes, placing no more than 500 µL in either collection receptacle.
18. Once the microtubes are filled, remove the PE-50 catheter from the femoral artery, ligate the proximal end, and close the incision with interrupted 5-0 prolene suture.
19. If blood is to be used later, place the caps on the filled pediatric EDTA tubes and store them at 4 °C.
20. If thrombus is to be prepared for same-day use, proceed to step 1.2 above.
21. Immediately repeat with the remaining centrifuge tubes.
22. Arterial blood donors may be returned to the vivarium following recovery from anesthesia if unilateral catheterization has been performed. The contralateral femoral artery could be catheterized later, but after bilateral femoral artery catheterization, donors should be euthanized in line with IACUC regulations.

5. TE-MCAo

NOTE: The TE-MCAo approach uses the standard cut down onto the common carotid artery. Choice of anesthesia and cut-down approach are up to the user. The protocol here starts at the point where the surgeon has already isolated the common carotid artery (CCA), the internal carotid artery (ICA) and the external carotid artery (ECA).

1. Expose the CCA and retract with a 3-0 silk ligature. Do the same for the ECA and the ICA.
2. Using a ligature or cautery, expose and occlude the Occipital artery.
3. Expose and retract or occlude the pterygopalatine artery (PPA).
   NOTE: With enough experience, this step can be omitted. We recommend always performing this step when learning the procedure.
4. Using gentle retraction, move the ECA and ICA into a "T" configuration.
5. Tie off the CCA ligature only after all of the above steps are completed.
6. Make an arteriotomy in the ECA (Zea Longa CCA open methodology¹⁰) using micro scissors.
7. Using the blunt-bevel method described in step 4.4.2, prepare the tip of one of the already loaded micro-catheters (inner diameter: 0.2 mm, outer diameter: 0.360 mm, length: 300 mm, connected to female Luer lock).
8. Carefully insert the microcatheter with pre-loaded thrombus into the ECA. Loosen the ICA restraining loop and advance the microcatheter intracranially to 16 mm at the carotid bifurcation under direct visualization to avoid inadvertent cannulation of the extracranial PPA.
   NOTE: If the catheter prefers to enter the CCA then gentle pressure with a pointed cotton tip applicator may help orient the advancing catheter into the ICA.
9. Secure the intraluminal microcatheter to ICA and CCA using 3-0 silk, being careful not to overtighten.
10. Inject 20 µL into the 1 mL syringe with a motorized syringe pump at a rate of 10 µL/min for a duration of 2 min.
    NOTE: Typically, the syringe pump will not advance at first until sufficient pressure builds up.
11. Inject an additional 20 µL of sterile saline using the same pump settings.
    NOTE: During the study of blood flow using laser speckled imaging, we found 100% success with a 20 µL flush.
12. Remove the microcatheter and permanently ligate the ECA with a suture proximal to the arteriotomy site. Remove all restraining loops. Observe carefully for any blood leakage and secure with additional ligatures if necessary.
13. Close the skin incision temporarily with staples or a running 5-0 prolene suture.
14. Remove the animal from anesthesia and place it in a warm recovery cage.
15. Optional: Labs may use laser Doppler flowmetry to confirm successful MCA occlusion. Alternatively, magnetic resonance imaging may be obtained at 24 or 48 h to confirm and quantify cerebral infarction.
    NOTE: Many users prefer to administer 1.5 mL of saline or lactated Ringer's solution subcutaneously at the conclusion of the surgery.

6. Systemic Thrombolysis

NOTE: Tenecteplase (TNK, 1.5 mg/mL) may be purchased for use as systemic thrombolysis. Users may prefer other lytic agents. Individual 50 mg vials are reconstituted with 50 mL of distilled water, and then 2.5 mL is precisely aliquoted into 3.0 mL vials. The aliquots may be frozen at -20 °C for later use. The stability of reconstituted recombinant tissue plasminogen activator stored at -20 °C is 60 days¹¹.

1. Thaw one frozen aliquot of TNK for each subject. Thaw at room temperature for about 15 min.
2. Using a 1 cc syringe, draw exactly 1.5 mg/kg. For example, a 300 g rat will need 0.45 mg.
3. Re-anesthetize the animal and remove the temporary skin closure.
4. Expose any branch of the external jugular vein.
   NOTE: The tail vein could be used if the laboratory is more comfortable with this route. We preferred using the jugular vein since SPAN surgeons are more comfortable with external jugular vein mobilization, and this is done routinely in our filament model.
5. Using a 30 or 32 G needle, deliver the TNK by direct puncture into the vein.
   NOTE: At first, users may wish to isolate the vein between 3-0 silk loops. With practice, this will not be necessary.
6. Close the incision with interrupted 5-0 prolene or another non-absorbable suture.
7. Return the animal to the recovery cage as in step 5.14.

7. Statistical methods

1. For making descriptive statistics, use mean and standard deviation for normally distributed numerical data, and use median and interquartile range for non-normally distributed data. Present categorical data as frequency distributions. For hypothesis testing, apply the t-test or ANOVA as appropriate for normally distributed parametric data; otherwise, do not perform hypothesis testing.

Results

Figure 1: Blood collection from donor animals and thrombus preparation. Blood was drawn from the donor animal through femoral artery catheterization and transferred to a microcentrifuge tube with (can be stored up to 4 weeks) or without EDTA (same day usage). To prepare the thrombus, calcium chloride solution was added to the stored blood tubes and immediately aspirated into a PE-50 catheter. The coiled PE-50 tubing was incubated in pre-warmed phosphate-buffered saline (PBS) at 37°C for 2 h in a table-top oven. After 2 h, blood-filled PE-50 tubing was immediately transferred to 4°C for storage. See section 1 in the protocol for more information. Please click here to view a larger version of this figure.

Figure 2: Loading thrombus into the catheter for injection. (A) On the day of MCAo surgery, thrombus was extruded from the PE-50 tubing into a petri dish containing PBS. (B) On the petri dish, the thrombus was trimmed to approximately 6 cm in length with a razor blade. (C) Thrombus was washed with PBS by drawing it into a PE-50 tubing and expelling it 5 times and then into a PE-10 tubing 15 times. (D) Thrombus was transferred to a petri dish containing concentrated Evan's Blue solution. (E) Thrombus was transferred to a petri dish containing diluted Evan's Blue solution and loaded into a prefabricated microcatheter. (F) The microcatheter was advanced through the internal carotid artery of the rat until 16 mm from the tip of the catheter passed from the carotid bifurcation of the common carotid artery and injected using a motorized syringe pump. See section 2 in the protocol for more information. Please click here to view a larger version of this figure.

In a pilot feasibility study, we enrolled 135 subjects at 6 research laboratories. Although subjects were randomized to one of 8 treatments, we report here on aggregate data only to illustrate the feasibility of the model. Of 135 enrolled subjects, TNK was given at the correct dose and time in 132 (98%). MRI was obtained 3 days after TE-MCAo in 102 (75%) due to animal death before scan in 33 (24%). Animal loss before Day 3 was consistent across all 6 laboratories (Table 1), indicating that the model can be deployed across multiple sites. The mean ± SD size of the lesions was 13 ± 16% (lesion area includes ischemia and edema as a percentage of the ipsilateral hemisphere) with some variability across laboratories (Table 2). Representative MR images and post-mortem emboli may be seen in Supplemental Figure 1.

	Laboratory
Variable	AG	DK	IW	MG	SD	YL
Sample Size (N)	23	30	11	28	24	19
MRI Done	19	24	9	23	14	13

Table 1: Feasibility by Laboratory. The study laboratories were Augusta University (AG), Duke (DK), University of Iowa (IW), Massachusetts General Hospital (MG), University of San Diego (SD) and Yale University (YL). The number of enrolled animals (N) and the number that survived to have MRI scanning was done Day 3 after stroke are shown. This data supports the feasibility of the model.

Site	Sample Size (N)	Lesion Volume (%) Mean ± SD	Midline Shift Index (%) Mean ± SD
AG	19	11 ± 14	11 ± 12
DK	24	18 ± 17	19 ± 13
IW	9	08 ± 13	13 ± 12
MG	23	09 ± 11	15 ± 17
SD	13	14 ± 21	04 ± 19
YL	12	18 ± 22	14 ± 15

Table 2: Volume of lesion seen on Day 3 MRI scans, by site. To establish the reproducibility of the model across six study sites, the lesion volume (ischemia plus edema) and midline shift were compared. Lesion volume is expressed as the areal percent of the ipsilateral hemisphere. Midline shift is expressed as the linear percent of the maximum width of the intracranial space. The data confirms heterogeneity across the laboratories.

Supplemental Figure 1: Representative Images. (A-E) Various multi-echo, multi-scan (MEMS) images of rat brains taken 48 h after TE-MCAo. The sections are approximately 1mm anterior to bregma. (F-J) Several examples of emboli lodged at the MCA origin. Please click here to download this figure.

Discussion

We refined the thromboembolic rodent model of ischemic stroke for use in a multi-laboratory pre-clinical testing network, SPAN. We greatly reduced the number of animals used by developing a method to store donor blood for later use. We also simplified the preparation of thromboemboli and the surgical approach to facilitate the performance of several MCAo procedures in one day. Our intention is that the TE-MCAo reach the same levels of user comfort and surgical volumes as are possible with the widely used nylon filament version of the MCAo. To achieve this result, we held two in-person surgeon training workshops, and all of the surgeons involved in this study had already several months, if not years, of experience with the more standard filament model. We demonstrated feasibility across six different research laboratories and showed excellent protocol adherence.

In choosing to implement this protocol, an investigator should consider the following items that determine protocol success. First, scrupulous care in collecting blood will influence successful thrombus formation. Arterial, not venous, blood should be collected to avoid premature clotting. Second, it is important to collect more blood than seems to be needed. Sometimes one specimen will fail to thrombose overnight, while the others will. Third, we strongly recommend that catheter loading be done using optical magnification, either surgical loupes with a headlamp or a magnifying lens with illumination. Fourth, cautious and careful dissection of the CCA, ICA, and ECA is essential to avoid tearing and intravascular thrombosis. Also, it is essential to visualize the PPA to ensure that the delivery catheter properly enters the ICA. Fifth, it is rare to see any adverse effect in the rodents due to the injection of thrombolytics. In our studies, we did not see any adverse events, such as angioedema or peripheral bleeding, following TNK injection. This is not to say it's not possible, but we are not aware of any reports of systemic side effects of thrombolytics in rodents. An investigator choosing to implement this protocol for the first time may well be advised to consult with one of the investigators named as co-author here, as we found personalized instruction to be very helpful.

In selecting an animal disease model for testing candidate therapies, it is important to accurately represent the patient population the drug is intended to treat. Focal ischemic stroke models include Rose Bengal photothrombotic models, cortical pial artery occlusion, and stereotactic endothelin-1 injection^13,14,15. These models have the advantage of creating highly reproducible lesions in predictable areas of the cortex. This advantage makes them preferable in studies of cellular mechanisms, or in correlating lesion location with behavioral outcomes. Added cost, complexity, and lack of reperfusion component make them less useful for drug screening.

The MCAo with intraluminal monofilament or thromboembolism are the established models of choice for large vessel occlusion (LVO)^10,16,17. The silicon-coated monofilament model is simple to execute and widely used. Although the location and size of the lesion are less predictable, in large-scale drug testing, this heterogeneity is viewed as an advantage because human strokes are also quite heterogenous^18,19. A disadvantage of the monofilament method is that the occlusion is biologically inert. In patients, thrombi and thrombolysis produce an inflammatory cascade associated with the release of thrombin, plasminogen, and other thrombus elements^20,21,22. To overcome this limitation, many have proposed models of thromboembolism in rabbit, porcine, canine, and non-human primate studies^{23,24,25,26,27}. These studies are complex and require specialized expertise in general. Nevertheless, we sought to adapt the thromboembolic model for use in our multi-laboratory, pre-clinical testing network. This required that we simplify the thromboemboli preparation, streamline the surgical approach, and significantly reduce the number of animals needed. We devised a thromboembolic model, the TE-MCAo, that could be deployed successfully at six different laboratories with excellent protocol adherence (Table 1) and reasonably consistent lesion size (Table 2). Throughput was sufficient, averaging 6 subjects per week.

There are several troubleshooting steps that may become useful. Failure of blood to thrombose is the most frequent source of protocol failure and is most often due to contamination of the preparation catheter with EDTA or heparin if placed on the surgical table. We recommend a scrupulous collection technique. We also recommend preparing the calcium chloride solution used to inactivate the EDTA fresh every week or every other week. After eluting the thrombus from the preparation catheter, it can be challenging to slice the thrombus into the required 5 cm fragments. Some thrombi are fragile and easily fractured. We recommend preparing the clots with optic magnification, and in extreme cases, moving on to a different donor thrombus. Aspirating and eluting thrombi requires a moderate level of skill; hence we recommend quite a long training period. Assure that the tip of the aspiration catheter is free of snags using optical magnification. Thrombi can be jammed in the implantation catheter - we recommend gentle 'back-and-forth' alternation between aspiration and elution. For surgeons experienced with the typical MCAo filament model, identification of the PPA may be a new step. We strongly recommend labeling the thrombi with Evan's blue, so that the catheter can be traced and followed as it ascends through the ICA. The bifurcation of the ICA/PPA may be easily located by careful, deep dissection following the ICA, and is reliably located deep to a thin branch of the internal carotid nerve.

This protocol comes with some limitations. There is an element of surgical skill necessary to perform cannulation of rodent carotid arteries. However, this is equally true of the filament model. Handling of the blood thrombi also requires surgical skill. As with all thromboembolic models, there is a lower rate of successful infarction compared to the filament model. If the goal of the investigation is more frequent incidents of infarction, the filament model would be preferable. On the other hand, if the goal is a replication of thrombus and thrombolysis then this model is desirable.

This model is intended for testing candidate cerebral protectants using a model that faithfully employs thrombus and thrombolysis. Future directions could include as well, studies of adjuvant therapy targeting the no reflow phenomenon. Many have proposed anticoagulants and antiplatelet agents for this purpose.

In conclusion, the TE-MCAo protocol presented here is adapted to allow deployment in multiple laboratories. We simplified the preparation of thrombi and streamlined the surgical approach. We devised a method to store donor blood, significantly reducing the numbers of animals needed.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
1 ml syringe	BD	309628
10x loupes
1cc syringe	Terumo	SS-01T
23g blunt needle	Grainger	5FVK7
3.0 cc syringe	BD	309656
3-0 silk ligature
30g needle	BD	305128
5-0 prolene suture	Ethicon	8698	To close the skin cut
5-0 silk ligature	Ethicon	A182
Aneurysm clamp
Atropine sulfate	Vet One	NDC86136-006-10	Reduce bronchial secretions
Blunt grooved forceps
Bupivacaine	Xelia Pharmaceuticals	NDC70594-118-01	Local anesthetics
Calcium chloride	Sigma	C7902
Carprofen	Zoetis Inc	NADA #141-199	Analgesic
Centrifuge tubes	Eppendorf	22364111	For blood collection
Ethanol	Koptec	UN1170
Evan’s Blue	Sigma	E2129
Heating pad	Harvard Apparatus	K021-435
Hemostatic clamp
Isoflurane	Vet One	NDC13985-528-60
Made By Me Ultimate Weaving loops	Horizon Groups	Not applicable	Soft restraints
Micro scissor	F.S.T.	15000-00
Microcatheters	Doccol Corporation	PE-FL-360-200-300	Prefabricated injection catheters
PE-10 catheter	BD	427401
PE-100 tubing	Intramedic	7426
PE-50 tubing	BD	427410
Pediatric EDTA tubes	BD Microtainer	K2E-363706	Store donor blood
Petri dish	Eppendorf	30702115
Phosphate buffered saline	Fisher	B2438
Prevantics	PDI Inc.	NDC10189-1080-1	Antiseptic
Q-tip	VWR	89031-288
Razor blade	AccuTec Pro	71960
Saline	Hospira, Inc.	NDC 0409-4888-02
Scissor	F.S.T.	14958-11
Sharp forceps
Syringe pump	New Era Pumps Inc.	BS-300
Tabletop oven	Boekel	132000
Tenecteplase (TNK)	Genentech, San Francisco	NDC50242-120-47	thrombolysis
Ear tag	RapID Lab, San Francisco, CA	RapID Tags