Improved reperfusion following alternative surgical approach for experimental stroke in mice

Background: Following ischemic stroke, recanalisation and restoration of blood flow to the affected area of the brain is critical and directly correlates with patient recovery. In vivo models of ischemic stroke show high variability in outcomes, which may be due to variability in reperfusion. We previously reported that a surgical refinement in the middle cerebral artery occlusion (MCAO) model of stroke, via repair of the common carotid artery (CCA), removes the reliance on the Circle of Willis for reperfusion and reduced infarct variability. Here we further assess this refined surgical approach on reperfusion characteristics following transient MCAO in mice. Methods: Mice underwent 60 min of MCAO, followed by either CCA repair or ligation at reperfusion. All mice underwent laser speckle contrast imaging at baseline, 24 h and 48 h post-MCAO. Results: CCA ligation reduced cerebral perfusion in the ipsilateral hemisphere compared to baseline (102.3 ± 4.57%) at 24 h (85.13 ± 16.09%; P < 0.01) and 48 h (75.04 ± 12.954%; P < 0.001) post-MCAO. Repair of the CCA returned perfusion to baseline (94.152 ± 2.44%) levels and perfusion was significantly improved compared to CCA ligation at both 24 h (102.83 ± 8.41%; P < 0.05) and 48 h (102.13 ± 9.34%; P < 0.001) post-MCAO. Conclusions: Our findings show CCA repair, an alternative surgical approach for MCAO, results in improved ischemic hemisphere perfusion during the acute phase.


Introduction
Cerebral ischemic stroke is one of the leading causes of death worldwide, carrying a significant disease burden as the secondleading cause of disability-adjusted life years 1,2 . Current approved treatment options are limited for acute ischemic stroke patients, involving either thrombolytic clot breakdown or physical clot removal. Recombinant tissue plasminogen activator (rtPA) is the only pharmacological treatment available with proven efficacy 3,4 , eligibility for this treatment is low and, due to a narrow therapeutic time window (<4.5 h), only ~1 5% of stroke patients are able to receive iv-rtPA treatment, with recanalisation success at less than 50% 5,6 . In addition to this, endovascular thrombectomy is increasingly being used in the treatment of large vessel occlusions, where the clot is removed from within the vessel allowing recanalisation, especially in patients who respond poorly to rtPA treatment 5 . Prompt restoration of blood flow to the ischemic zone is the primary clinical goal, potentially salvaging tissue and preventing the spread of ischemic damage 7 . Recanalization of occluded vessels is positively correlated with improved survival rates and recovery outcomes for ischemic stroke patients 8 .
To allow development of new clinical therapeutics for stroke, the pathophysiology and mechanisms of disease/recovery need to be elucidated. Preclinical stroke models that closely mimic mechanisms of injury (and recovery) allow investigation of potential clinically viable treatments. Although experimental stroke models have shown the positive benefits of various potential therapeutics, these have repeatedly failed during clinical trials, preventing their progression into the clinic. This continued lack of translation requires us to further refine preclinical studies and to allow the evolution of more representative models 9,10 . Stroke in the territory of the middle cerebral artery is the most common presentation seen in patients 20 , thus targeting this vessel in preclinical models is valid. However, clot lysis induced by rtPA treatment results in gradual reperfusion to the ischemic territory, whereas in the MCAO model, filament removal is sudden giving an immediate blood flow surge to the ischemic area 21 . In the clinic, endovascular thrombectomy allows physical removal of large vessel clots 21,22 , permitting rapid reflow through the previously occluded vessel, which is the primary clinical goal and parallels that which occurs in transient MCAO upon filament withdrawal. However, MCAO models where intraluminal access is obtained via the CCA, such as the Koizumi method 23 , rely on collateral perfusion after filament removal due to the CCA being permanently tied off to prevent blood loss. Previously, we reported a modification to the Koizumi model of intraluminal filament MCAO, which allows bilateral CCA reperfusion, negating the reliance on the CoW 10 . In commonly used mice strains such as the C57BL/6J, the variability seen in lesion volume may be largely due to anatomical variations in the CoW 17 . Removing reperfusion reliance on the CoW reduces the variability seen in lesion volume but the impact on reperfusion is not known. Our aim here therefore is to determine if the adapted MCAO model can improve reperfusion following removal of the filament.

Animals
This study was conducted in accordance with the UK Animals (Scientific Procedures) Act, 1986, under Project license P28AA2253 following ethical approval from the University of Manchester. Animals were group housed (n=4) on arrival in temperature-and humidity-controlled individually ventilated mouse cages (IVCs) with a 12-h light/dark cycle, situated within a specific pathogen free (SPF) facility. Cages contained: woodchip bedding, paper sizzle nesting and cardboard tube. Animals were given adlib access to dry pellet food and water. Experimental sample size was determined using a power calculation from preliminary data (one tailed, α=0.001, β=0.9), taking into account a mortality rate of 10% established from historical MCAO data within our group. A total of 18 adult male C57BL/6J mice (10 weeks old on arrival; Charles River, UK), weighing 22-33 g at the time of MCAO, were used during this study: two animals were excluded due to surgical complications, two due to reaching humane endpoints and one due to repeated wound complications. All remaining animals (n=13) were humanely killed by cervical dislocation under anaesthesia following the 48-h post-MCAO scanning procedure. All experiments are reported in accordance with the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines 2018 24 .

Study design
Individual animals were randomly assigned using an automated list randomiser (random.org) to one of two experimental groups: CCA ligation or CCA repair. The CCA ligation group

Amendments from Version 2
Following reviewer comments the concluding paragraph has been rewritten to improve clarity and sentence structure.
Any further responses from the reviewers can be found at the end of the article REVISED underwent 60 minutes transient MCAO with permanent CCA ligation following reperfusion 25 . The CCA repair group underwent 60 minutes transient MCAO followed by repair of the CCA using a tissue pad and sealant combination, resulting in a patent CCA 10 . Both groups underwent laser speckle contrast imaging (LSCI). The order of coded LSCI raw data files was randomised prior to analysis using an automated list randomiser. The experimenter was blind to groups during LSCI scanning and image analysis for raw data, data was decoded for statistical analysis only.
Laser speckle contrast imaging All animals underwent LSCI. Imaging was undertaken 12 days prior to MCAO (baseline imaging), then at 24 h and 48 h post-MCAO ( Figure 1A). Anaesthesia was induced and maintained with isoflurane (induction 5% in 100% O 2 ; maintenance 1.5% in N 2 O/O 2 , 70/30%), following which mice were placed in a stereotaxic frame on a heat mat set to 37°C. The animals were positioned underneath a moorFLPI 2 Full-Field Perfusion Imager (Moor Instruments, UK). Prior to incision Bupivacaine hydrochloride (2 mg/kg, Marcain 0.25%, AstraZeneca, UK) local anaesthesia was injected subcutaneous around the intended incision site. A midline incision was made and skin retracted to expose the intact skull. Ultrasound gel was applied to the skull, followed by a 10-mm 2 glass coverslip pressed down slightly to allow complete coverage of the intended scan area. LSCI was conducted for 3 min at 19 frames (20 ms per frame, 10-s intervals).
The incision was cleaned free of ultrasound gel using sterile swabs and flushed with sterile saline, the incision was sutured using subcutaneous 6-0 dissolvable sutures followed by tissue adhesive application (Vetbond, US). The animal was returned to a pre-warmed clean cage for recovery. Following baseline LSCI only, animals were given access to paracetamol (200 mg/kg/24 h; Calpol ® , Johnson & Johnson Ltd UK) in jelly for 24 h for self-administration analgesia. LSCI images were analysed using moorFLPI2 Full-Field Laser Perfusion Imager Review v5.0 software. For each animal, ipsilateral and contralateral hemisphere ROIs were drawn, ROI flux data was obtained from each hemisphere for each of the 19 images per scan, the mean of these values was used in later analyses. Ipsilateral hemisphere cerebral blood flow (CBF) is expressed as a percentage of the contralateral hemisphere CBF, within each imaging session. All analysis was conducted blinded to experimental conditions.

Surgical procedure
All mice underwent transient MCAO as previously detailed, and surgical methods used were identical to those published previously by the author 10,26 , modified only to exclude Laser Doppler Flowmetry recording at MCAO reperfusion. MCAO was confirmed using laser doppler flowmetry (LDF), distinguished as a significant sustained drop in LDF flux once the microfilament filament was advanced into the internal carotid artery.

Statistical analysis
All data are expressed as mean ± standard deviation of the mean (SD). All statistical analysis was performed using Prism v8 for windows (GraphPad Software, CA). The criterion for statistical significance was P ≤ 0.05. Distribution of data was assessed using Shapiro-Wilk normality test, prior to statistical analysis. Repeated measure two-way analysis of variance (ANOVA) test was used to determine effect of time and intervention, repair and ligation. Post-hoc Sidak corrected multiple comparison testing was then used to compare within groups against baseline/pre-MCAO, and within time point between groups.

Discussion
Here we report CCA vessel repair following MCAO in mice results in improved reperfusion of the ischemic hemisphere, assessed using LSCI. Ligation of the CCA following MCAO filament removal significantly reduced perfusion to the ischemic hemisphere for a prolonged period post-surgery and the CBF did The data indicate CCA repair improves ischemic area perfusion post-MCAO, we assume this is due to the return to patency of the CCA removing the reliance on collateral flow through the CoW. The CoW has been shown in C57BL/6J mice to be anatomically variable across individual animals 17 , adding variability to collateral supply to the ischemic area. Similarly, Smith et al.
reported, using MCA territory laser doppler flowmetry to measure perfusion, significantly increased post-MCAO perfusion using the Longa ECA entry method with perfusion returning to comparable baseline levels, whereas use of the Koizumi CCA entry method showed only a return to 50% of baseline levels 23 .
The key difference between the two methods is that the Longa method allows reperfusion through the bilateral CCAs whereas the Koizumi method relies on collateral supply to perfuse the ischemic area due to only the CCA being patent. These results mimic those reported here with CCA repair resulting in bilaterally patent CCAs leading to a return to baseline in perfusion post-MCAO. The use of a muscle pad combined with a fibrin sealant was previously shown in rat MCAO to allow forward blood flow through the CCA with complete occlusion of the vessel not present following the procedure 27 . Additionally, the CCA repair method ensures that the ECA remains in situ and patent, reducing the impact on animal welfare and recovery 19 .
Allowing reperfusion to occur through all available vessels transforms the MCAO model to be more representative of clinical endovascular thrombectomy. Endovascular thrombectomy is employed in an increasing number of clinical cases of large vessel occlusion (LVO), allowing mechanical, within vessel, removal of the occluding clot leading to a surge reperfusion of the ischemic zone, such as is seen following MCAO filament removal 21 . There is discussion in the literature as to the relevance of MCAO for modelling clinical stroke, due to the slow gradual clot dispersion that occurs following rtPA treatment, a mechanism unlike the sudden increase in flow seen at MCAO filament removal 28 , with suggestion that this slower clinical rtPA treatment reperfusion profile is closer to other embolic stroke models 21,29 . However, due to the positive effects of using endovascular thrombectomy for ischemic stroke treatment, on blood flow and patient outcome, the intervention is likely to become the primary treatment for LVO 21 . With this in mind, there is renewed interest in the intraluminal filament model of MCAO as an established model of endovascular thrombectomy.
The data presented here suggest a beneficial effect on post-MCAO reperfusion in the ischemic hemisphere however, the limitations of our approach must be considered. LSCI has very good spatial and temporal resolution and is used to monitor CBF changes under both physiological and pathophysiological conditions but it may be that the low penetration depth of the laser does not permit measurements of deep brain CBF but only allows imaging of surface and pial vessels 30 . This limits the interpretation of the data presented here, in terms of deep level reperfusion and further investigation would be required to assess the ability of CCA repair to improve deep level perfusion. In addition to this limitation, cerebral perfusion data is typically reported as raw arbitrary units or normalised to a baseline recording 31,32 , whereas the data reported here is presented as a % of the contralateral or non-ischemic hemisphere. Often, MCAO experiments undertaking LSCI utilise continuous imaging pre and post-MCAO, preventing variability in animal/camera placement, focus level and instrument settings, improving the validity of obtained data and allowing within experiment data normalisation/analysis. Due to the experimental design of this study, data was normalised within experiment to the non-stroke hemisphere, as reported previously 33-36 , to reduce the impact of between experiment variability from, for example equipment setup. However, it is important to note that this approach does not take into account the dynamic transhemispheric effect on CBF shown during focal ischemia reperfusion 37 . In addition to the expected CBF variations during and following MCAO, CBF in the contralateral hemisphere has also been reported to be affected during this pathology. The contralateral hemisphere is an important reserve to balance blood flow along the ischemic gradient to the ipsilateral hemisphere via collateral flow, particularly where reperfusion is reliant on flow from the basilar artery and contralateral CCA 18 . It is also worthy to note that a sustained reduction in CBF measurements within the CCA ligated group is clear. This hypoperfusion could lead to exacerbation of poststroke outcomes, with reduced oxygen levels resulting in pathological changes such as inflammation, demyelination and apoptosis 38 . Chronic sub-ischemic hypoperfusion is a key characteristic of conditions such as vascular dementia and there is a strong clinical association between hypoperfusion and development of vascular dementia 39 . In a further study it would be interesting to assess the impact of improved reperfusion on brain injury/recovery and functional outcome associated with ischemia.
In summary, this brief report shows that CCA repair following MCAO, in the mouse, using a homologous tissue pad combined with sealant improves blood flow to the ischemic hemisphere. The results validate the use of CCA repair in MCAO studies to remove reliance on the CoW avoiding the variation associated with this structure, in a commonly used mouse strain, on post-stroke measures. The outcome reported here may be of particular interest for use in endovascular thrombectomy research, mimicking the clinical situation of clot removal without ligation of arteries.
This project contains the following underlying data: • BASE_Repair (baseline raw LCSI export files for animals that underwent repair of the CCA; the three-symbol code at the start of each file represents a different mouse).
• 24h_Repair (24-hour raw LCSI export files for animals that underwent repair of the CCA; the three-symbol code at the start of each file represents a different mouse).
• 48h_Repair (48-hour raw LCSI export files for animals that underwent repair of the CCA; the three-symbol code at the start of each file represents a different mouse).
• BASE_Ligated (baseline raw LCSI export files for animals that underwent ligation of the CCA; the threesymbol code at the start of each file represents a different mouse).
• 24h_Ligated (24-hour raw LCSI export files for animals that underwent ligation of the CCA; the threesymbol code at the start of each file represents a different mouse).
• 48h_Ligated (48-hour raw LCSI export files for animals that underwent ligation of the CCA; the threesymbol code at the start of each file represents a different mouse).

4.
Thank you to the authors for their detailed response. The manuscript has improved and my comments have all been adequately addressed. The only minor point is the request to change the second last sentence of the discussion "Validating the use of CCA repair..." does not seem to have changed, and this sentence still reads as an incomplete sentence. However, this is very minor and should not prevent acceptance.
I co-authored a paper in 2017 with both Stuart Allan and Claire Gibson who are Competing Interests: authors on the submitted manuscript. The paper is here https://journals.sagepub.com/doi/10.1177/0271678X17709185 and was part of large consortium of researchers in the pre-clinical stroke field putting together best practice guidelines. However, this did not influence my ability to provide an unbiased review.
Reviewer Expertise: Stroke and cerebral blood flow I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. 1.

5.
6. reduced infarct volume size and variability. This manuscript provides a nice addition with Laser Speckle Contrast Imaging showing reperfusion changes lasting 48 hours between the traditional Koizumi method and the CCA repair method. This simple repair method could improve our ability to conduct stroke modelling. I have some comments below.
Specific comments: The Laser Speckle Contrast Imaging method has great spatial and temporal resolution but does not provide any depth assessment of CBF changes in the brain. It only provides a measurement of CBF on the surface of the brain. Therefore, all that can be stated from this study is that pial CBF can be improved post-MCA during CCA repair compared to the CCA ligated method, as we do not know from this imaging technique whether the underlying brain is fully reperfused. Some discussion around this would be useful.
The Laser Speckle Contrast Imaging was conducted at baseline (prior to MCAO), 24h post-MCAO and 48h post-MCAO. However, there was no apparent assessment during MCAO to show that a stroke had actually occurred. How can you confirm that these animals actually had a stroke with no CBF assessment during MCAO?
The sustained (48h) changes shown in Figure 1 are interesting. It would have been excellent to have shown how these CBF values are associated with changes in neurological deficit (could be done by a neuroscore) or infarct volume at 48 hours. Were the brains assessed for this?
The CCA ligated results show sustained hypoperfusion on the surface of the brain for 48 hours. Some extended discussion on how this could impact the development of the lesion and chronic injury would be useful, particularly since chronic occlusion of the CCA is used in vascular dementia studies with chronic neurodegeneration and impaired cognition present. 1.

Is the study design appropriate and is the work technically sound? Yes
Are sufficient details of methods and analysis provided to allow replication by others? Yes

If applicable, is the statistical analysis and its interpretation appropriate? Yes
Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Yes I co-authored a paper in 2017 with both Stuart Allan and Claire Gibson who are Competing Interests: authors on the submitted manuscript. The paper is here https://journals.sagepub.com/doi/10.1177/0271678X17709185 and was part of large consortium of researchers in the pre-clinical stroke field putting together best practice guidelines. However, this did not influence my ability to provide an unbiased review.
Reviewer Expertise: Stroke and cerebral blood flow I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
Author Response 29 Apr 2020 , University of Nottingham, Nottingham, UK claire gibson The Laser Speckle Contrast Imaging method has great spatial and temporal resolution but does not provide any depth assessment of CBF changes in the brain. It only provides a measurement of CBF on the surface of the brain. Therefore, all that can be stated from this study is that pial CBF can be improved post-MCA during CCA repair compared to the CCA ligated method, as we do not know from this imaging technique whether the underlying brain is fully reperfused. Some discussion around this would be useful. Response: Please see added section into discussion.
The Laser Speckle Contrast Imaging was conducted at baseline (prior to MCAO), 24h post-MCAO and 48h post-MCAO. However, there was no apparent assessment during MCAO to show that a stroke had actually occurred. How can you confirm that these animals actually had a stroke with no CBF assessment during MCAO? Response: As now stated in the methods section of the paper the drop in CBF during MCAO was confirmed using laser doppler flowmetry.
The sustained (48h) changes shown in Figure 1 are interesting. It would have been 1.

1.
The sustained (48h) changes shown in Figure 1 are interesting. It would have been excellent to have shown how these CBF values are associated with changes in neurological deficit (could be done by a neuroscore) or infarct volume at 48 hours. Were the brains assessed for this? Response: We would agree that any further analysis under these experimental conditions would be useful. However, that was beyond the remit of this current study which was powered to answer the specific research question. We have previously reported the effect of this modified surgical approach on lesion volume (see Trotman et  and others). Given that the quantification of the ipsilateral CBF is a % of the contralateral, any changes in CBF in the contralateral hemisphere in response to MCAO or reperfusion could alter the reference point for this analysis. Some discussion should be added about this as a limitation. Response: Please see amended section in discussion.