New Grading System Based on Magnetic Resonance Imaging in a Mouse Model of Subarachnoid Hemorrhage
Background and Purpose—A grading system for experimental subarachnoid hemorrhage (SAH) that does not require animal euthanasia is currently unavailable. We proposed a new grading system based on MRI and evaluated the feasibility of this method in a mouse model of SAH.
Methods—SAH was induced by endovascular perforation in adult male C57BL/6 mice. Mice underwent MRI 24 hours after SAH and were categorized into the following 5 grades based on T2*-weighted imaging: Grade 0, no visible SAH or intraventricular hemorrhage (IVH); Grade 1, minimal/localized SAH without IVH; Grade 2, minimal/localized SAH with IVH; Grade 3, thick/diffuse SAH without IVH; and Grade 4, thick/diffuse SAH with IVH. Neurological deficits were then assessed and the mice euthanized for conventional SAH grading.
Results—Among a total of 47 mice, 4% were scored as grade 0, 30% as grade 1, 11% as grade 2, 30% as grade 3, and 26% as grade 4. This MRI grading had excellent interobserver reliability (weighted κ value =0.94), and there were strong correlations between the MRI grading and the conventional grading (r=0.85; P<0.001) or between MRI grade and neurological scores (r=−0.46; P<0.01).
Conclusions—The new MRI grading correlated well with conventional grading and enabled in vivo evaluation of SAH severity. This grading system may offer advantages in future studies of experimental SAH.
The number of publications using rodent models of subarachnoid hemorrhage (SAH) has been rapidly increasing.1 The endovascular perforation method for inducing SAH is frequently used in rodents and is considered to the closest imitation of human SAH.2,3 In that model, there are greater inter-animal variations in the extent and volume of subarachnoid clot compared with other SAH models.4 A method established by Sugawara et al5 is commonly used to evaluate the SAH severity in perforation models of rodent SAH, but this is performed after euthanasia by visualizing clot extension and thickness at the base of the brain. A grading system that does not require animal euthanasia is currently unavailable. In this study, we developed a new SAH grading system for the mouse endovascular perforation model based on MRI.
Materials and Methods
All animal protocols were approved by the University of Michigan Committee on the Use and Care of Animals. A total of 58 male C57BL/6 mice (Charles River Laboratories, Portage, MI) weighing 22 to 30 g were used. Food and water were available to all animals ad libitum before and after surgery. SAH was induced by single surgeon (Y. Egashira) using endovascular perforation technique as previously described.6 MRI was performed 24 hours after SAH using a 7.0-T Varian MR scanner (Varian Inc, Palo Alto, CA) with acquisition of T2 fast spin-echo and T2* gradient-echo sequences with a field of view of 20×20 mm, matrix of 256×256 mm, and 25 coronal slices (0.5 mm thick).7 We categorized animals into the following 5 grades according to the thickness of SAH clot and the presence of intraventricular hemorrhage (IVH) in lateral ventricle detected by T2*-weighted imaging (T2*WI): grade 0 indicates no visible SAH or IVH; grade 1 indicates minimal/localized SAH with no IVH; grade 2 indicates minimal/localized SAH with IVH; grade 3 indicates thick/diffuse SAH (hematoma exceeding 0.5 mm thickness is visible in >2 slices of T2*WI) with no IVH; and grade 4 indicates thick/diffuse SAH with IVH. Neurological scores were also evaluated at 24 hours after SAH by a blinded observer as previously described.8 Mice were given a score of 3 to 18 in 1-number steps (higher scores indicate greater function).
Mice were then euthanized and the brains examined. The severity of SAH was assessed using a conventional grading system.5 The basal brain, including brain stem, was divided into 6 segments. Each segment was assigned a grade from 0 to 3 depending on the amount of SAH as follows: 0, no SAH; 1, minimal SAH; 2, moderate SAH with recognizable arteries; and 3, SAH covering the cerebral arteries. The animals were assigned a total score ranging from 0 to 18 by summing the scores from all 6 segments.
Data in text are expressed as the mean±SD. Commercially available software (JMP 7; SAS Institute Inc, Cary, NC) was used for all statistical analysis. The Cohen κ (Kc) and weighted κ (Kw) values were calculated to examine reproducibility of the scale between 2 investigators.9 The correlation between MRI grade and either conventional SAH severity score or neurological score was examined by Spearman’s rank correlation test. P<0.05 values were considered statistically significant.
The mortality rate was 19% (11 of 58) at 24 hours after SAH induction. Among a total of 47 remaining mice, 2 (4%) were scored as grade 0 by MRI grading system, 14 (30%) were grade 1, 5 (11%) were grade 2, 14 (30%) were grade 3, and 12 (26%) were grade 4 (Table). Kw value for the thickness of the subarachnoid clot and Kc value for the presence of IVH in the lateral ventricle were 0.94 and 0.82, respectively (Tables I and II in the online-only Data Supplement). Kw value for the MRI grading scale was 0.94 (Table III in the online-only Data Supplement). Thus, it was considered that this system has excellent reproducibility. Figure 1 shows the representative T2- and T2*-weighted MR images and corresponding postmortem photos of basal brain in each MRI grade. The conventional SAH severity scores were 3.5±0.7 in grade 0, 6.4±1.4 in grade 1, 8.6±1.8 in grade 2, 10.5±2.0 in grade 3, and 12.8±1.9 in grade 4. There was a strong correlation between MRI grade and conventional SAH score (r=0.85; P<0.001; Spearman’s rank correlation test; Figure 2A). There was also a significant correlation between MRI grade and neurological score (r=−0.46; P<0.01; Spearman’s rank correlation test; Figure 2B). The neurological scores were 13.5±3.5 in grade 0, 14.0±2.8 in grade 1, 11.4±2.4 in grade 2, 12.8±3.0 in grade 3, and 9.4±2.8 in grade 4. As expected, there was also a significant correlation between SAH severity score and neurological score (r = -0.39; P<0.01; Spearman’s rank correlation test; data not shown).
In the present study, we developed a new reproducible SAH grading system based on T2*-weighted MRI. In endovascular perforation models of SAH in rodents, brain injury is related to the SAH bleeding (severity).5 In addition, it has been reported that SAH-related conditions, such as acute hydrocephalus and white matter injury, are also affected by SAH severity.7,10 It is expected that SAH severity may also relate to delayed SAH-induced complications, such as chronic hydrocephalus, as in humans. The MRI grading system enables in vivo evaluation of SAH severity without animal euthanasia and might offer advantages especially in long-term experimental studies. Thus, for example, it may be used to (a) examine the extent to which SAH severity affects long-term outcome, (b) reduce the model variability in therapeutic trials by ensuring that animals have a similar initial SAH severity, and (c) determine whether the effects of therapeutic interventions vary with SAH severity.
Our scale imitated the human SAH scale using computed tomography established by Claassen et al.11 Interestingly, the proportion of each grade in mice is comparable to that in human. Mice with IVH in the lateral ventricle tended to have worse neurological scores than those without IVH. In humans, SAH with IVH also tends to lead to worse clinical outcomes.11 MRI makes it possible to detect IVH, and the effect of IVH in SAH-induced brain injury should be determined in future experimental studies.
Although the present study has used T2*-weighted images to examine degree and location of SAH, the utility of other forms of imaging (susceptibility weighted imaging, quantitative susceptibility mapping) merits investigation. T2-weighted MR images can also be used to examine SAH-induced injury, allowing the relationship between T2WI abnormalities and SAH grading to be examined in future studies.
Our new MRI grading system correlated well with conventional grading system and had excellent reproducibility. This system may offer advantages in future studies using rodent SAH models.
Sources of Funding
This study was supported by grants NS-073595, NS-079157, and NS-084049 from National Institutes of Health, 973 Program–2014CB541600, and Japan Heart Foundation/Bayer Yakuhin Research Grant Abroad.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.114.007834/-/DC1.
- Received October 17, 2014.
- Revision received December 2, 2014.
- Accepted December 4, 2014.
- © 2014 American Heart Association, Inc.
- Kamii H,
- Kato I,
- Kinouchi H,
- Chan PH,
- Epstein CJ,
- Akabane A,
- et al
- Egashira Y,
- Hua Y,
- Keep RF,
- Xi G
- Okubo S,
- Strahle J,
- Keep RF,
- Hua Y,
- Xi G
- Claassen J,
- Bernardini GL,
- Kreiter K,
- Bates J,
- Du YE,
- Copeland D,
- et al