Journal of Clinical Neuroscience 16 (2009) 545–548

Contents lists available at ScienceDirect

Journal of Clinical Neuroscience

journal homepage: www.elsevier .com/ locate/ jocn

Laboratory Study

A double-injection model of intracerebral hemorrhage in rabbits

Zhen Yu, Li-Fen Chen, Xiao-Feng Li, Dong-Ping Zhang, Yang-Mei Chen, Wan-Fu Wu, Chang-Lin Hu *

Department of Neurology, The Second Affiliated Hospital, Chongqing University of Medical Sciences, 76 Lin Jiang Men, Chongqing 400010, China

a r t i c l e i n f o a b s t r a c t

Article history:Received 9 February 2008Accepted 27 April 2008

Keywords:Animal modelBrain edemaIntracerebral hemorrhageRabbits

0967-5868/$ - see front matter � 2008 Elsevier Ltd. Adoi:10.1016/j.jocn.2008.04.026

* Corresponding author. Fax: +86 23 67038978.E-mail address: huclcq@gmail.com (C.-L. Hu).

We aimed to develop a double-injection model of intracerebral hemorrhage (ICH) in rabbits and to eval-uate it as a tool for investigating post-ICH brain injury. Rabbits were injected with 300 lL fresh autolo-gous whole blood into the right basal ganglia. Behavioral changes were rated, brain water content (BWC)was measured and brain tissue morphology was also examined. ICH was established in 93.5% of the bloodinjection group. At 1, 3 and 7 days after ICH, there were significant differences in the total neurologicalscores (p < 0.01) and BWC (p < 0.01) between a sham-operated group and the ICH group. These findingssuggest that the model produces a persistent neurological deficit, hematoma volume and perihematomaledema and closely mimics human hypertensive basal ganglia ICH; it is a controllable and reproduciblehematoma that lends itself to quantitative investigation.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Spontaneous intracerebral hemorrhage (ICH) constitutes 15% ofall strokes in the USA and Europe and 20% to 30% in Asian and blackpopulations. ICH is associated with high mortality and disability.1

No widely approved effective acute medical treatment exists,and surgical evacuation of ICH is not beneficial.2 To improve theclinical outcome of ICH, we need to better understand the patho-genesis of ICH-induced brain injury. Appropriate animal modelsmay assist in elucidating the mechanisms of brain injury after ICH.

Two models have principally been used to mimic spontaneousICH. One uses bacterial collagenase to induce ICH, and the otheris an autologous whole blood model. Bacterial collagenase digeststhe collagen present in the basal lamina of blood vessels and causesbleeding into the surrounding brain tissue.3 Although the collage-nase method is a simple and reproducible way to produce hemor-rhage, bacterial collagenase causes significant inflammation andlikely differs from the mechanism that produces ICH in humans.4

The autologous whole blood model has been widely developed inrats.5–7 However, this method produces hematomas of varying sizebecause of ventricular rupture and the backflow of infused bloodalong the needle track, which leads to intraventricular and/or sub-arachnoid blood leakage.8 Subsequently, Deinsberger et al. first re-ported a double-injection model of ICH in rats,9 which appeared tobe a good approximation of the occurrence of ICH. The model wasrepeatedly validated in rats and mice.10,11 and is usually used toinvestigate mechanisms of hemorrhagic damage.12 Therefore, therodent double-injection model has become a popular model inICH research.

ll rights reserved.

Primate models probably yield results that are, in general, mostsimilar to human ICH events. Other desirable features of an animalmodel include availability, ease in handling, and low cost. Althoughdogs and cats have been used in ICH models, their companion ani-mal status and cost make them less attractive. Rat models involveblood infusions in the striatum and have focused solely on graymatter injury.13 Hence, considering all aspects and types of ICHmodels, the rabbit model should be the best choice. Because oftheir similarity in phylogeny to primates, rabbits possess compara-ble cerebrovascular morphological features that may be involvedin the pathophysiology and pathochemistry of human ICH.14,15

The rabbit model was used in ICH research as early as the1950s.16 Our knowledge of experimental ICH in rabbits is largelylimited to the single injection model.17–21 However, hematomasize is not reproducible, and therefore the single injection modelin rabbits is not suitable for quantitative research of ICH. A similarproblem is seen in the rodent ICH model.8,17,22

We conducted the present study to develop a double-injectionmodel in rabbits and to assess neurological deficit, brain watercontent (BWC), and brain tissue morphology.

2. Materials and methods

2.1. Animal preparation and groups

The animal protocol for our study was approved by the Institu-tional Animal Care and Use Committee of Chongqing University ofMedical Sciences. Adult male New Zealand white rabbits (n = 76)weighed 3.1 kg to 3.7 kg and were 5–6 months old.

Animals were randomly divided into two groups: the sham-operated group (Sham, n = 30) and the ICH group (ICH, n = 46).ICH rabbits were injected with 300 lL of blood into the basal

546 Z. Yu et al. / Journal of Clinical Neuroscience 16 (2009) 545–548

ganglia. Sham animals were subjected to the same manipulationsas ICH rabbits, but no blood was injected.

2.2. The ICH model

ICH was produced by a modified double-injection method, sim-ilar to that described.23

1. Anesthesia, fixation and orientation: the rabbit was anesthe-tized with an abdominal cavity injection of 10% chloral hydrate(500 mg/kg) and then placed in a stereotactic frame (David KopfInstruments, Tujunga, CA, USA). The scalp was incised longitu-dinally in the midline, and a 2 mm burr hole was made (2mm anterior to the bregma, 6 mm lateral to the midline) inthe skull using a dental drill. The double cannula (made in ourlaboratory) was composed of a 6 mm long outer cannula (20-gauge cylindrical plastic cannula) and an 8.5 mm long innercannula (25-gauge stainless steel cannula with needle core).The double cannula was inserted into the right basal gangliawith stereotaxic guidance (2 mm anterior to the bregma, 6mm lateral to the midline, at a depth of 6 mm below the surfaceof the skull).

2. A two-step blood injection: autologous whole blood was takenfrom the right femoral artery, which was catheterized andattached to a micro-infusion pump (Braun, Melsungen, Ger-many). First, each rabbit was injected with 50 lL of fresh autol-ogous whole blood for 2 min (at 25 lL/min) and then the needlecore was inserted into the inner cannula and left in situ for7 min. The inner cannula was slowly inserted 2.5 mm furtherdownwards and the needle core was pulled out from the innercannula. Then 250 lL blood was injected over 6 min using amicroinfusion pump (at 40 lL/min), and the needle core wasreinserted into the inner cannula.

3. Three needle withdrawals: after 7 min, the inner cannula wasslowly withdrawn upward 2.5 mm and then held in place for8 min, the double cannula was slowly withdrawn upward 3mm and then held in place for 9 min and, finally, the doublecannula was slowly removed. Sham animals were subjected tothe treatment as ICH rabbits, but no blood was injected. Theburr hole was filled with bone wax, and the incision in the skinwas closed with suture after infusion. The animal was placed inan incubator with free access to food and water. The body tem-perature was maintained at 37.0 ± 0.5�C using a warm waterblanket.

2.3. Behavioral testing

Behavior was assessed in 7 rabbits in each group at 1, 3, and7 days after ICH using a neurological deficit scale.24 The scaleincluded tests of motor function (1–4), consciousness (1–4), headturning (0–1), circling (0–1) and hemianopsia (0–1). A total scoreof 11 indicates maximum impairment (comatose or dead rabbit),whereas 2 denotes complete normality. Tests were conducted byan observer blinded to the two groups.

2.4. Brain water content

At 6 h and 1, 3 and 7 days after ICH, 7 animals in each group andtime-point were decapitated and brains were removed rapidly. Acoronal brain slice (3 mm thick) 5 mm from the frontal pole wastaken. We weighed the tissue samples immediately on an elec-tronic analytical balance (Changzhou Instruments, Changzhou,China) to obtain the wet weight. The brain samples were thendried at 110�C in an Electric Blast Drying Oven (Sida Apparatus,Chongqing, China) for 24 h to obtain the dry weight. We calculated

the water percentage as follows: (wet weight � dry weight)/wetweight � 100%.

2.5. Histological examination

1. Tissue fixation: at 6 h and 1, 3 and 7 days after ICH, the brains oftwo animals were perfusion fixed with a mixture of 4% parafor-maldehyde, glacial acetic acid, and methanol (1:1:8 by volume)for at least 7 days and then cut into 3 mm thick coronal slices.Under macroscopic view we noted the effects of compression,herniation, and ventricular extension, and then embedded thebrain blocks in paraffin. Sections (10 lm thick) were cut inthe coronal plane and stained with hematoxylin and eosin.

2. Measurement of hematoma volume: at 1 h after ICH, the brainwas quickly removed, placed in a �20�C refrigerator for 8 minand sectioned coronally at 2 mm intervals. The hematoma areawas defined with the software TD2000 image analysis system(Tiandi-bainian, Beijing, China). We then calculated hematomavolume by multiplying the measured area by the thickness ofthe section interval. The hematoma volume was expressed inmicrolitres (lL).

2.6. Statistical analysis

All data in this study were presented as mean ± standard errorof the mean (SEM). All behavioral and histologic analyses were per-formed by experimenters who were blind to group identity. Datawere analyzed with the Student’s t-test or analysis of variance withthe Scheffé F test. Differences were considered to be statisticallysignificant at p < 0.05.

3. Results

In this study, all rabbits tolerated the surgical procedure welland there was no surgical mortality. Three animals were excludedfrom the study due to blood rupturing into the ventricular systemor subarachnoid space. ICH was established in 93.5% of ICH grouprabbits (43/46) without rupturing into the ventricular system orsubarachnoid or subdural spaces. Histological examination of thebrain revealed a localized hematoma in all animals with a bloodinfusion (Fig. 1). Hematoma volume of the ICH group at 1 h was246.73 ± 35.56 lL (82.33%). In contrast, the sham-operated groupshowed only a small, non-hemorrhagic lesion. We found perihe-matoma edema in the white matter 6 h after ICH. The mean ±SEM BWC for the infused rabbits at 1 day was 82.43 ± 0.87% andthis value increased significantly to 83.95 ± 0.58% at 3 days and de-creased to 79.01 ± 0.75% at 7 days. There were significant differ-ences in BWC between the two groups at 1, 3, and 7 days(p < 0.001, p < 0.001, p = 0.002, respectively) (Fig. 2). We observedpersistent left hemiparesis after blood injection but not after can-nula insertion alone. The total neurologic score (mean ± SEM) forthe infused rabbits at 1 day was 8.1 ± 1.2, and this improved signif-icantly to 4.6 ± 1.7 at 3 days and 2.8 ± 0.7 at 7 days. There was asignificant difference in the total neurologic scores between theICH and sham groups at 1, 3, and 7 days (p < 0.001, p = 0.001,p = 0.004, respectively) (Fig. 3).

4. Discussion

We developed a basal ganglia ICH model in rabbits based on amodified double-injection method. The infusion of 300 lL of bloodinduced marked and reproducible brain edema and neurologicdeficits.

We modified several aspects of the double-injection method ofDeinsberger et al.9 First, our double-injection ICH model in rabbits

Fig. 1. Double-injection model of ICH in the rabbit. (A) sequential coronal sections of the brain, (B) sagittal section of the brain. Moderate-sized basal ganglion hematoma isevident. Blood does not rupture into the ventricular system. This figure is available in colour at www.sciencedirect.com.

0

2

4

6

8

10

0 1 3 7

Time after ICH (day)

Tota

l Neu

rolo

gic

Scor

es

Sham

Blood

*

* *

*

Fig. 3. The time course of neurologic deficits after intracerebral hemorrhage (ICH)in rabbits. Total neurologic score (normal score 2; maximal score 11) was measuredbefore ICH and at 1, 3, 7 days after ICH between blood injection group (blood) andsham-operated group(sham). Values are expressed as mean ± SEM; n = 7 for eachgroup, *P < 0.01, **P < 0.001 vs. sham group.

76

78

80

82

84

86

6 24 72 168Time after ICH (hour)

brai

n w

ater

con

tent

(%)

Sham

Blood* *

*

* *

Fig. 2. Brain water content of ipsilateral hemisphere after intracerebral hemor-rhage (ICH) in rabbits at 6, 24, 72, 168 hours between the 300 lL blood injectiongroup (blood) and the sham-operated group (sham). Values are expressed as mean± SEM; n = 7 for each group. �P < 0.01, ��P < 0.001 vs. sham group.

Z. Yu et al. / Journal of Clinical Neuroscience 16 (2009) 545–548 547

using a double cannula is different to the guide cannula used inother autologous blood ICH models. The inner cannula is movablewithin the outer cannula, allowing its clear orientation in thehematoma. Thus, the double cannula is not only an easier way toinfuse blood or a drug into the hematoma cavity, but it also helpsto block interstices around the needle track, involves less invasivesurgery, and effectively prevents blood regurgitation through theinjection canal into the subarachnoid space. Second, comparedwith other autologous blood ICH models, our ICH model incorpo-rates blood injection in two steps and three needle withdrawals.The meeting of the second needle withdrawal with negative pres-sure attracts unclotted blood, and clotted blood is initially infusedin the needle track because the enhanced blood coagulation effectof the clot completely blocks the needle track and prevents the lossof blood during needle withdrawal. Third, the speed and pressureof blood injection were controlled by a microinfusion pump. OurICH model utilized a slow infusion of unheparinized autologouswhole blood into the basal ganglia and limited its extravasation

into the subarachnoid or ventricular spaces; this more closelymimics the natural events that occur with spontaneous ICH in hu-mans. Furthermore, slow infusion also avoided undesired non-physiological pressure injury to the adjacent tissue. Fourth,although Belayev et al. found that the use of heparin facilitatedthe infusion of whole blood taken from the heart of a donormouse,11 the infusion of unheparinized blood produces signifi-cantly greater edema than heparinized blood,25 so we prefer touse unheparinized autologous whole blood.

This basal ganglia ICH model has clinical relevance for severalreasons: (i) the basal ganglia ICH is the most frequent hemorrhagesite in humans, (ii) basal ganglia ICH in rabbits resulted in moredamage to the striatum, white matter, and cortex than that in ratsor mice, (iii) white matter damage was related to the long-termoutcome after ICH.26 Thus, the model closely mimics humanhypertensive basal ganglia ICH.

548 Z. Yu et al. / Journal of Clinical Neuroscience 16 (2009) 545–548

The hematoma in the infused rabbits was consistently locatedin the outer area of the basal ganglia with focal extension intothe interior area of the cerebral cortex and the anterolateral inter-nal capsule (Fig. 1). All ICH in rabbits developed without refluxalong the needle track; only 6.5% (3/46) ruptured into the ventric-ular system. The hematoma was made of blood cells surroundingthe needle tract. It is difficult to translate data from rabbits to hu-mans. ICH volumes in humans are typically 100-fold larger thanthose in rabbits. Thus, a 300 lL ICH in rabbits would correspondto hematoma of about 30 or 40 mL in humans. Therefore, a300 lL ICH in rabbits could be considered of moderate size; itcan impact adjacent tissue surrounding the hematoma, but it doesnot result in brain herniation and death. It produces a highly repro-ducible hematoma that lends itself to quantitative investigation.

Perihematomal edema develops immediately after an ICH andpeaks several days later.27,28 Edema development after ICH can ele-vate intracranial pressure and cause herniation, brain stem com-pression, and death.29,30 In this study, perihematoma edemacould be found in the white matter at 6 h after ICH, and BWC pro-gressively increased during the first 24 h and peaked at 3 days,then declined gradually (Fig. 2). This is similar to observations inother models.25,31,32 There were also more severe neurological def-icits in the ICH group at 1 and 3 days after ICH and these weremaintained during 1 week of monitoring (Fig. 3). The worseningneurological deficits in the ICH group were associated with en-hanced brain edema; perihematomal edema was more severe inthe ICH group compared to that in the sham-operated group at 3days after ICH but not at 1 day. These results suggested that brainedema is significant in determining brain injury after ICH. The roleof perihematomal edema in clinical deterioration is controversialand needs further investigation.13

Within 1 day of injection of autologous blood into a rabbit brain,there was persistent contralateral hemiparesis or hemiplegia afterblood injection but not after cannula insertion alone. The scoringsystem appeared sensitive to detecting deficits after unilateralbrain injury in the rabbits. Behavioral testing is important not onlyin clinical stroke patients but also in animal models of ICH.11 A bat-tery of behavioral tests for assessing acute changes needs to bedeveloped for the rabbit model of ICH.

5. Conclusions

The rabbit model produces a consistent neurological deficit,hematoma volume and brain edema. This model closely mimicshuman hypertensive basal ganglia ICH. It produces a controllableand reproducible hematoma that lends itself to quantitative mea-surement. Therefore, the double-injection model in rabbits canbe used to investigate the mechanisms of and therapeutic inter-ventions for ICH.

Acknowledgment

We gratefully acknowledge the technical assistance of Dr Wen-bing Wu.

References

1. Qureshi A, Tuhrim S, Broderick J, et al. Spontaneous intracerebral hemorrhage.N Engl J Med 2001;344:1450–60.

2. Mendelow AD, Gregson BA, Fernandes HM, et al. Early surgery versus initialconservative treatment in patients with spontaneous supratentorialintracerebral haematomas in the International Surgical Trial in IntracerebralHaemorrhage (STICH): a randomised trial. Lancet 2005;365:387–97.

3. Rosenberg GA, Mun-Bryce S, Wesley M, et al. Collagenase induced intracerebralhemorrhage in rats. Stroke 1990;21:801–7.

4. Del Bigio MR, Yan HJ, Buist R, et al. Experimental intracerebral hemorrhage inrats: magnetic resonance imaging and histopathological correlates. Stroke.1996;27:2312–9.

5. Laurent JP, Molinari GF, Oakley JC. Experimental model of intracerebralhematoma. J Neuropathol Exp Neurol 1976;35:560–8.

6. Bullock R, Mendelow AD, Teasdale GM, et al. Intracranial haemorrhage inducedat arterial pressure in the rat. Part 1: Description of technique, ICP changesandneuropathological findings. Neurol Res 1984;6:184–8.

7. Nath FP, Jenkins A, Mendelow AD, et al. Early hemodynamic changes inexperimental intracerebral hemorrhage. J Neurosurg 1986;65:697–703.

8. Yang GY, Betz AL, Chenevert TL, et al. Experimental intracerebral hemorrhage:relationship between brain edema, blood flow, and blood-brain barrierpermeability in rats. J Neurosurg 1994;81:93–102.

9. Deinsberger W, Vogel J, Kuschinsky W, et al. Experimental intracerebralhemorrhage: description of a double injection model in rats. Neurol Res1996;18:475–7.

10. Hickenbottom SL, Grotta JC, Strong R, et al. Nuclear factor-jB and cell deathafter experimental intracerebral hemorrhage in rats. Stroke 1999;30:2472–7.

11. Belayev L, Saul I, Curbelo K, et al. Experimental intracerebral hemorrhage in themouse:histological, behavioral, and hemodynamic characterization of doubleinjection model. Stroke 2003;34:2221–7.

12. Xi G, Keep RF, Hoff JT. Mechanisms of brain injury after intracerebralhaemorrhage. Lancet Neurol 2006;5:53–63.

13. NINDS ICH Workshop Participants. Priorities for clinical research inintracerebral hemorrhage. Stroke 2005; 36: 23-41.

14. Eldevik OP, Kristiansen K, Torvik A. Subarachnoid hemorrhage andcerebrovascular spasm. Morphological study of intracranial arteries based onanimal experiments and human autopsies. J Neurosurg 1981;55:869–76.

15. Tanabe Y, Sakata K, Yamada H, et al. Cerebral vasospasm and ultrastructuralchanges in cerebral arterial wall. An experimental study. J Neurosurg1978;49:229–38.

16. Sawyer CH, Everett JW, Green JD. The rabbit diencephalon in stereotaxiccoordinates. J Comp Neurol 1954;101:801–24.

17. Kaufman HH, Pruessner JL, Bernstein DP, et al. A rabbit model of intracerebralhematoma. Acta Neuropathol (Berl) 1985;65:318–21.

18. Narayan RK, Narayan TM, Katz DA, et al. Lysis of intracranial hematomas withurokinase in a rabbit model. J Neurosurg 1985;624:580–6.

19. Gustaffson O, Rossitti S, Ericsson A, et al. MR imaging of experimentallyinduced intracranial hemorrhage in rabbits during the first 6 hours. Acta Radiol1999;40:360–8.

20. Qureshi AI, Ling GS, Khan J, et al. Quantitative analysis of injured, necrotic, andapoptotic cells in a new experimental model of intracerebral hemorrhage. CritCare Med 2001;29:152–7.

21. Koeppen AH, Dickson AC, Smith J. Heme oxygenase in experimentalintracerebral hemorrhage: the benefitof tin-mesoporphyrin. J Neuropathol ExpNeurol 2004;63:587–97.

22. Andaluz N, Zuccarello M, Wagner KR. Experimental animal models ofintracerebral hemorrhage. Neurosurg Clin N Am 2002;13:385–93.

23. Zhang Y, Zhang Y, Chen X, et al. Development of stable and reproducibleintracerebral hematoma model in rabbits. Chinese J Exp Surg 2001;18:269–71.

24. Purdy PD, Devous Sr MD, Batjer HH, et al. Microfibrillar collagen model ofcanine cerebral infarction. Stroke 1989;20:1361–7.

25. Xi G, Keep RF, Hoff JT. Erythrocytes and delayed brain edema formationfollowing intracerebral hemorrhage in rats. J Neurosurg 1998;89:991–6.

26. Wagner KR, Xi G, Hua Y, et al. Lobar intracerebral hemorrhage model in pigs:rapid edema development in perihematomal white matter. Stroke 1996;27:490–7.

27. Xi G, Keep RF, Hoff JT. Pathophysiology of brain edema formation. NeurosurgClin N Am 2002;13:371–83.

28. Suzuki J, Ebina T. Sequential changes in tissue surrounding ICH. In: Pia HW,Longmaid C, Zierski J, editors. Spontaneous Intracerebral Hematomas. Berlin:Springer; 1980. p. 121–8.

29. Diringer MN. Intracerebral hemorrhage: pathophysiology and management.Crit Care Med 1993;21:1591–603.

30. Ropper AH. Lateral displacement of the brain and level of consciousness inpatients with an acute hemispheral mass. N Engl J Med 1986;314:953–8.

31. Enzmann DR, Britt RH, Lyons BE, et al. Natural history of experimentalintracerebral hemorrhage: sonography, computed tomography andneuropathology. AJNR Am J Neuroradiol 1981;2:517–26.

32. Hua Y, Schallert T, Keep RF, et al. Behavioral tests after intracerebralhemorrhage in the rat. Stroke 2002;33:2478–84.